Application of mitochondrial differentiation expression gene in preparation of depressive disorder diagnostic agent and therapeutic drug

Abstract

The invention discloses application of a mitochondrial differentiation expression gene in preparation of a depression diagnostic agent and a depression treatment drug, and relates to the technical field of depression diagnosis and treatment. And the mitochondrial differentiation expression gene comprises one or more of TFAM (Transforming Factor Amplification Molecular), SURF1 and NDUFB9. According to the present invention, the SURF1 gene simultaneously expressed in inhibitory neurons and excitatory neurons is focused on, the key effect of the SURF1 in CUMS-induced depression-like behaviors is found, the CUMS-induced molecular and behavior changes are effectively reversed by the overexpression of the SURF1, and the potential anti-depression effect is displayed; besides, animal experiments find that overexpression of the SURF1 can effectively recover the function of an electron transport chain, ATP generation is enhanced, mitochondrial membrane potential is maintained, and oxidative stress is reduced. The effects jointly relieve depression-like behaviors induced by CUMS, and further support the protection effect of the SURF1 in a depression model, which provides evidence for the new effect of the SURF1 in the pathophysiology of depression.

Description

Technical Field

[0001] The present invention relates to the technical field of diagnosis and treatment of depression, and in particular to the application of mitochondrial differentially expressed genes in the preparation of diagnostic agents and therapeutic drugs for depression. Background Art

[0002] Major depressive disorder (MDD) is a common psychiatric disorder characterized by chronically depressed mood and impaired cognitive function, which, in severe cases, can lead to self-harm or suicidal tendencies. To date, depression affects over 300 million people worldwide. Despite extensive research on the biological mechanisms of depression, its precise pathophysiological processes remain incompletely understood. Currently, the diagnosis and treatment of depression rely primarily on subjective symptom assessments, lacking reliable objective clinical biomarkers. Furthermore, existing antidepressant medications often have limited efficacy and are associated with numerous side effects, imposing significant physical, psychological, and economic burdens on patients. Studies have shown that individuals with chronic sleep disorders are more likely to develop mental health problems, including bipolar disorder, suicidal thoughts, and, in particular, depression. Importantly, a bidirectional relationship may exist between MDD and sleep disorders (SD): Depressed individuals are prone to sleep disturbances due to abnormal neurotransmitters, and chronic sleep problems further exacerbate depressive symptoms. Therefore, exploring the shared pathological pathways between depression and sleep disorders is crucial for a deeper understanding of the pathophysiological basis of depression.

[0003] Recent studies have revealed a close relationship between mitochondrial dysfunction and the pathophysiology of sleep disorders and depression. Mitochondria, as essential organelles in eukaryotic cells, participate in multiple key physiological processes, such as ATP synthesis, the generation and clearance of reactive oxygen species (ROS), and the regulation of apoptosis during the cell cycle. Studies have shown that mitochondrial dysfunction is closely associated with the development and progression of depression, potentially exacerbating depressive symptoms by affecting neuronal energy metabolism, altering ROS levels, releasing inflammatory factors, and interfering with neurotransmitter synthesis and signaling. Furthermore, mitochondria play a crucial role in regulating the sleep-wake cycle. During sleep, mitochondria maintain normal brain function by redistributing energy and clearing metabolic waste. However, mitochondrial dysfunction may lead to insufficient energy supply, disrupting neural networks, and triggering or exacerbating sleep disorders. Despite these findings, the specific regulatory mechanisms of mitochondrial dysfunction in depression remain incompletely understood and urgently require further investigation. Summary of the Invention

[0004] In response to the above problems, the present invention aims to provide the use of mitochondrial differentially expressed genes in the preparation of diagnostic agents and therapeutic drugs for depression.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] In one aspect, the present invention provides the use of differentially expressed mitochondrial genes in the preparation of diagnostic agents and therapeutic drugs for depression, wherein the differentially expressed mitochondrial genes include one or more of TFAM, SURF1 and NDUFB9.

[0007] Specifically, the mitochondrial differentially expressed gene is SURF1.

[0008] Specifically, SURF1 mRNA expression is decreased in depression.

[0009] On the other hand, the present invention also provides a research method for the application as described above, comprising the following steps:

[0010] S1: Identification of differentially expressed mitochondrial genes associated with depression using transcriptomics and single-cell transcriptomics sequencing technologies;

[0011] S2: The relationship between SURF1 overexpression and CUMS-induced depressive-like behavior in mice was verified by sucrose preference test, forced swim test, and open field test.

[0012] S3: Analysis of mitochondrial function in mice by overexpression of SURF1 and restoration of mitochondrial function induced by CUMS;

[0013] S4: Further validate the protective effect of SURF1 against mitochondrial dysfunction caused by CUMS using targeted inhibitors.

[0014] On the other hand, the present invention also provides use of a reagent for detecting the SURF1 gene in preparing a diagnostic agent for depression.

[0015] On the other hand, the present invention also provides use of a SURF1 gene activator in preparing a drug for treating depression.

[0016] On the other hand, the present invention also provides the use of a product containing SURF1 protein in the preparation of a drug for treating depression.

[0017] The beneficial effects of the present invention are:

[0018] 1. Through bioinformatics analysis, the present invention identified key differentially expressed genes associated with depression: TFAM, SURF1, and NDUFB9. Further focusing on the SURF1 gene, which is expressed in both inhibitory and excitatory neurons, the authors found that SURF1 plays a key role in CUMS-induced depressive-like behavior. Overexpression of SURF1 effectively reversed the molecular and behavioral changes induced by CUMS, demonstrating a potential antidepressant effect. This provides evidence for a new role for SURF1 in the pathophysiology of major depressive disorder (MDD), contributes to a deeper understanding of the pathophysiological basis of MDD, and provides a potential therapeutic target for the treatment of MDD associated with sleep disorders (SD).

[0019] 2. The present invention also investigated the effect of SURF1 overexpression on restoring mitochondrial function in CUMS mice. The results showed that SURF1 overexpression effectively restored electron transport chain function, enhanced ATP production, maintained mitochondrial membrane potential, and reduced oxidative stress. These effects collectively alleviated CUMS-induced depressive-like behaviors, further supporting the protective role of SURF1 in this depressive model. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is the identification of differentially expressed genes in the present invention.

[0021] Among them, A, volcano plot of differentially expressed genes: the x-axis represents log2 fold change (FC), and the y-axis represents -log10 (adj.p.val); B, Venn diagram of intersection genes: the bar chart on the left shows the number of genes in each subset, and the bar chart on the top shows the number of intersection genes; C, PPI network; D, Venn diagram of cytoHubba.

[0022] Figure 2 This is the identification of the hub gene in the present invention.

[0023] Among them, A, LASSO regression results: the x-axis represents the log(lambda) value, and the y-axis represents the degrees of freedom; B, the relationship between the generalization error of SVM and the number of features; C, Boruta gene importance ranking: green represents important features, yellow represents uncertain importance, and red represents unimportant features. The blue box shows the minimum, mean, and maximum Z-score of the shaded attribute; D, Venn diagram of hub genes; E and F, expression patterns of hub genes between different groups: the box color represents the sample group, and the x-axis color represents gene regulation (up-regulation is purple, down-regulation is green, and no significant change is black); the top annotation indicates the significance level: ns indicates p>0.05, *p≤0.05, **p≤0.01, ***p≤0.001, ****p≤0.0001; G and H, ROC curves of hub genes: the x-axis represents the false positive rate, and lower x-values ​​indicate higher accuracy; the y-axis represents the true positive rate, and higher y-values ​​indicate higher accuracy.

[0024] Figure 3 This is the identification of key cell types and hub gene expression profiles in single-cell transcriptomes in the present invention.

[0025] Among them: A, UMAP cell cluster map; B, cell annotation bubble map; C, UMAP cell cluster annotation map; D and E, biomarker expression profiles; F, cell communication results. The size of the circle represents the number of cells; larger circles represent higher cell numbers. Cells with arrows express ligands, and arrows point to cells expressing receptors. In the left panel, thicker lines indicate more ligand-receptor pairs. In the right panel, higher interaction probability / strength (sum of probabilities) indicates thicker lines.

[0026] Figure 4 The overexpression of SURF1 in the present invention alleviates the depressive-like behavior of mice induced by CUMS.

[0027] AC, mRNA levels of TFAM, SURF1, and NDUFB9; D and E, protein levels of SURF1; F and G, immunofluorescence results showing the localization and expression of SURF1, scale bar = 100 μm; H, percentage of sucrose preference over 6 weeks; I, immobility time in the FST; J, total distance traveled in the OFT; K, number of crossings of the central area in the OFT; L and M, TUNEL assay results revealing neuronal apoptosis in the PFC, scale bar = 100 μm. Data are expressed as mean ± standard error (n = 6 per group). *p < 0.05 and **p < 0.01 indicate significance compared with the control group (Vector). #p < 0.05 and ##p < 0.01 indicate significance compared with the CUMS (Vector) group.

[0028] Figure 5In the present invention, overexpression of SURF1 rescued mitochondrial abnormalities in the PFC of CUMS mice.

[0029] Figure 5: A, G, protein levels of NDUFS1, SDHB, UQCRFS1 / RISP, COX5B, and ATP5A in the PFC; H, measurement of ATP production in the PFC; I and J, assessment of mitochondrial membrane potential using flow cytometry; K and L, measurement of ROS production in the PFC using the DHE assay. Scale bar = 100 μm. Data are presented as mean ± standard error (n = 6 per group). *p < 0.05 and **p < 0.01 indicate significant differences compared with the control group (vector). #p < 0.05 and ##p < 0.01 indicate significant differences compared with the CUMS (vector) group.

[0030] Figure 6 In the present invention, overexpression of SURF1 restored mitochondrial biogenesis and mitophagy in the PFC of CUMS mice.

[0031] A, D, protein levels of PCG-1α, NRF1, and TFAM in the PFC; E and F, immunofluorescence results show colocalization of LC3 and TOMM20; scale bar = 100 μm; G, protein levels of PINK1 and Parkin in the PFC; J, protein levels of p62, Beclin, and LC3 in the PFC. Data are expressed as mean ± standard error (n = 6 per group). *p < 0.05 and **p < 0.01 indicate significant differences compared with the control group (vector). #p < 0.05 and ##p < 0.01 indicate significant differences compared with the CUMS group (vector).

[0032] Figure 7 In the present invention, 3-MA or Mdivi-1 inhibited the protective effect of SURF1 on mitochondrial dysfunction caused by CUMS.

[0033] A and B, immunofluorescence results showing colocalization of LC3 and TOMM20, scale bar = 100 μm; CI, protein levels of NDUFS1, SDHB, UQCRFS1 / RISP, COX5B, and ATP5A in the prefrontal cortex (PFC); J, measurement of ATP production in the prefrontal cortex (PFC); K and L, ROS generation in the prefrontal cortex (PFC) was measured using a DHE kit, scale bar = 100 μm; M and N, mitochondrial membrane potential was assessed by flow cytometry. Data are presented as mean ± standard error of the mean (SEM) (n = 6 per group). *p < 0.05 and **p < 0.01 indicate significance compared with the SURF1-OE group. DETAILED DESCRIPTION

[0034] In order to enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0035] This study explored potential biomarkers of depression and its underlying mechanisms by conducting bioinformatics analysis of key differentially expressed genes in peripheral blood transcriptome data from patients with major depressive disorder, focusing on genes associated with mitochondrial dysfunction (MD-RGs) and genes associated with sleep disorders (SD-RGs). In addition, single-cell transcriptomics technology was used to explore key brain cells that affect patients with major depressive disorder. Finally, molecular biology experiments were used to verify the analysis results and further explore the role and mechanism of key genes in depression. This study specifically includes the following contents:

[0036] 1. Research Methods

[0037] 1.1 Data Collection

[0038] Transcriptome training set: Gene expression and associated phenotypic information were obtained from the GSE52790 dataset in the GEO database. The platform used was GPL17976, and the sequencing samples were derived from peripheral blood mononuclear cells. The training set included a total of 22 samples: 12 control samples (Control) and 10 samples with major depression (Case).

[0039] Transcriptome validation set: Gene expression and associated phenotypic information were obtained from the GSE38206 dataset in the GEO database. The platform used was GPL13607, and the sequencing samples were derived from peripheral blood mononuclear cells. The validation set included a total of 36 samples, including 18 control samples (Control) and 18 samples with major depression (Case).

[0040] Mononuclear cell RNA sequencing (scRNA-seq): We downloaded mononuclear cell sequencing data from the GSE144136 dataset from the GEO database. The samples were derived from the prefrontal cortex and included a total of 34 samples: 17 control samples (Control) and 17 major depression samples (Case).

[0041] Mitochondrial-related genes (MD-RGs): A total of 1,136 MD-RGs were downloaded from the MitoCarta3.0 database.

[0042] Sleep disorder-related genes (SD-RGs): A total of 2,891 SD-RGs were retrieved from GeneCards by searching the keyword sleep disorder and screening for a relevance score ≥5.

[0043] 1.2 Bioinformatics analysis

[0044] 1.2.1 Differential gene expression (DEG) analysis: Differential gene expression (DEG) analysis between the control group (Control) and the disease group (Case) was performed using the "limma" package in the R language. Data normalization and batch effect correction were then performed to ensure the accuracy of the results. A linear model was applied to detect differentially expressed genes, with an adj.p. value threshold of <0.05.

[0045] 1.2.2 Functional enrichment analysis: To evaluate the functions of differentially expressed genes (DEGs), this study used two widely used databases: GO and KEGG. Using the clusterProfiler package in R, DEGs were mapped to GO categories for annotation, systematically exploring their roles inside and outside the cell. This study also mapped genes to the KEGG pathway database to identify potential signaling pathways and metabolic pathways. To assess enrichment significance, hypergeometric tests or Fisher's exact tests were applied to detect overrepresentation of DEGs in specific categories and pathways. To control false positives, the Benjamini-Hochberg method was applied for multiple testing correction, and significant enrichment terms with an FDR < 0.05 were screened.

[0046] 1.2.3 Protein-Protein Interaction (PPI) Network: A PPI network was constructed using the STRING database to identify key proteins associated with differentially expressed genes and their predicted interactions. This approach helps elucidate the functional relationships between proteins and their roles in disease processes. The PPI network was visualized using Cytoscape software, and topological analysis was performed to identify central nodes and key pathways involved in disease progression.

[0047] 1.2.4 Machine learning for gene selection: To optimize the candidate gene list and improve the research accuracy, this study applied three machine learning techniques for feature selection: least absolute shrinkage and selection operator (LASSO), support vector machine (SVM) and Boruta algorithm.

[0048] (1) LASSO regression: LASSO is a regularization technique that introduces a penalty term to identify genes that are strongly associated with the disease and reduce redundant variables, thereby improving the generalization ability of the model and avoiding overfitting.

[0049] (2) SVM: SVM classifies genes that are critical for disease development based on their gene expression levels, ensuring that the selected subset is highly predictive.

[0050] (3) Boruta algorithm: Boruta, based on random forest, eliminates irrelevant features and identifies important genes by evaluating their importance with randomly generated features, which is very suitable for high-dimensional data.

[0051] 1.2.5 ROC Curve Analysis: In this study, ROC curve analysis was performed using the "pROC" package in the R language to evaluate the potential of selected genes as diagnostic and prognostic biomarkers. The diagnostic accuracy of candidate genes was assessed using the area under the curve (AUC), which quantifies predictive accuracy. AUC values ​​range from 0 to 1, with values ​​closer to 1 indicating high sensitivity and specificity, meaning that the gene can effectively distinguish between diseased and healthy individuals.

[0052] 1.2.6 Single-cell RNA Sequencing Analysis: This study used the Seurat package in the R language to process clustered single-cell RNA sequencing data. Seurat provides tools for preprocessing, quality control, dimensionality reduction, clustering, and identification of differentially expressed genes. First, the data were normalized and denoised to remove technical noise while preserving biological signals. Next, principal component analysis (PCA) was performed to reduce data complexity and identify underlying cell populations. Finally, cluster analysis was applied to identify the primary cell types expressing the target genes.

[0053] 1.3 Animal experiments and grouping

[0054] The animals used in this study were male C57BL / 6 mice. The mice were between 8 and 10 weeks old and weighed between 25 and 30 grams. The mice were randomly assigned to four groups to ensure comparability of the experimental data: 1) Control-Vehicle: mice that received only vehicle treatment and no stress exposure. 2) Control-SURF1-OE: mice that overexpressed the SURF1 gene. 3) CUMS-Vehicle: mice that were exposed to CUMS and then treated with vehicle (CUMS (Chronic Unpredictable Mild Stress) is a widely used method for studying depression in animal experiments). 4) CUMS-SURF1-OE: mice that were exposed to CUMS and overexpressed the SURF1 gene.

[0055] To further clarify the specific role of mitophagy in chronic unpredictable mild stress (CUMS), two targeted inhibitors were used in the study: 3-MA (HY-19312, MCE, 10 mg / kg / day) and Mdivi-1 (HY-15886, MCE, 3 mg / kg / day), both administered by intraperitoneal injection. Both inhibitors were administered from the fourth week after the start of CUMS for two weeks. The following subgroups were further established in the CUMS-SURF1-OE group: 1) CUMS-SURF1-OE group: mice overexpressing SURF1 under CUMS conditions and not receiving inhibitor treatment; 2) CUMS-SURF1-OE+3-MA group: mice treated with 3-MA; 3) CUMS-SURF1-OE+Mdivi-1 group: mice treated with Mdivi-1.

[0056] 1.4 Stereotaxic administration of drugs to the prefrontal cortex (PFC)

[0057] This study used adeno-associated virus (AAV) vectors provided by BrainVTA. During the injection process, the mice were first anesthetized. After anesthesia, the mice were placed in a stereotactic apparatus for precise brain positioning. The injection coordinates were determined according to the standard brain atlas as follows: anterior-posterior (AP) = +2.3 mm, mediolateral (ML) = ±0.3 mm, dorsoventral (DV) = -2.0 mm. A total of 500 nL of AAV vector was injected into each cortical area. To ensure stable delivery of the vector, the injection speed was set at 40 nL / minute. Three weeks after the injection, the CUMS model was successfully established.

[0058] 1.5 Chronic Unpredictable Stress Model (CUMS)

[0059] CUMS was used to induce depressive-like behavior in mice. To minimize external disturbances and ensure consistent environmental conditions, mice were housed individually in cages within a control room throughout the experiment. A series of stressors were randomly assigned to the animals, with one stressor applied daily. To maintain the unpredictability of stress exposure and prevent habituation, the specific stressor was randomly varied weekly for six consecutive weeks. Stressors included: 1) 1 hour of restraint; 2) 5 minutes of swimming in 4°C or 45°C water; 3) 24 hours of food or water deprivation; 4) 1 minute of tail clamping; 5) 12 hours of continuous nighttime lighting; 6) bedding moistened with 100 grams of sawdust absorbed in 200 milliliters of water; and 7) cage tilt at a 45° angle for 24 hours.

[0060] 1.6 Sucrose preference test (SPT)

[0061] The SPT is commonly used in animal behavioral studies to assess anhedonia—diminished pleasure in response to rewarding stimuli—under stress. During the training phase, mice are exposed to two bottles of 1% sucrose solution for 24 hours to familiarize themselves with the environment and the substance. During the acclimation phase, one of the sucrose bottles is replaced with water to discriminate between water and sweetness. Afterward, mice undergo 24 hours of water deprivation to induce thirst and enhance their motivation for fluid intake. The testing phase, which assesses anhedonia, involves providing two bottles: one containing sucrose solution and one containing water. To avoid spatial bias, the positions of the bottles are swapped every 12 hours. At the end of the test, sucrose preference is calculated by dividing sucrose consumption by total fluid intake and multiplying by 100%. Higher preference scores indicate greater motivation for the sucrose solution, while lower scores indicate anhedonia, which is often associated with stress-related mood changes.

[0062] 1.7 Forced swim test (FST)

[0063] FST is a widely used experimental method for assessing depressive-like behavior in mice. In this test, each mouse is placed individually in a clear Plexiglas cylinder filled with water, and the water temperature is maintained between 23-25°C. The test consists of an initial adaptation phase and a formal test phase. During the initial adaptation phase, the mice become familiar with the environment. During the formal test phase, the mice need to swim in the water for 6 minutes, followed by a 1-minute adaptation period. The next 5 minutes are used to measure the immobility time of the mice. Immobility is defined as the mouse remaining still, with only minimal movement to keep its head above the water or floating vertically. The entire process is recorded by video for subsequent analysis, and new water is added between each trial. By analyzing immobility behavior, researchers can assess the depressive-like behavior of mice.

[0064] 1.8 Open Field Test (OFT)

[0065] The OFT is used to assess physical activity and anxiety-like behavior in mice. In this experiment, each mouse is placed in the center of a square open field, and its behavior is continuously monitored and recorded by a SuperMaze V2.0 video system. During a five-minute free exploration period, the main measurements are the frequency with which the mouse enters the center area and the total distance walked. The frequency of entering the center area reflects the anxiety level of the mouse, as anxious mice tend to avoid the center and prefer to stay near the edges. On the other hand, the total distance walked indicates the level of physical activity; longer distances generally indicate lower anxiety levels and more exploratory behavior. All data are systematically recorded by the video system for subsequent analysis to evaluate the effects of different experimental conditions on mouse behavior.

[0066] 1.9qRT-PCR

[0067] This study first used TRIzol reagent to extract total RNA from mouse cortical tissue, and then reverse transcribed the extracted RNA into complementary DNA (cDNA) using Top GreenqPCR Kit. Next, cDNA was quantitatively analyzed using real-time quantitative PCR (qPCR), and Gapdh mRNA was used as an internal reference gene for normalization to account for sample differences. Data analysis used the 2-ΔΔCq method to determine the relative expression level of the target gene by calculating the difference in Cq values ​​between the target gene and the internal reference gene. Specifically, the ΔCq of each sample was first calculated (the Cq value of the target gene minus the Cq value of Gapdh), and then the ΔΔCq was determined by comparing the ΔCq values ​​between the experimental group and the control group. Finally, 2 (-ΔΔCq) The relative expression levels of target genes were calculated using the following formula. The detailed primer sequences used in this study are listed in Table 1.

[0068] Table 1 qRT-PCR primer sequences

[0069] RNA Forward Primer(5'to3') Reverse Primer(5' to 3') Tfam TAGGCACCGTATTGCGTGAG GTGCTTTrAGCACGCTCCAC Surf1 TTGCTCTGACTTGGGGAAGAA GGGCTTCCTGTTTTCTGTGAG Ndufb9 GAAGGTGCTGCGGCTGTAT TCTTCAAACCGGGCTCTCAT Mfn1 CTCTTTCGGGAGGATGAGACAG CAGCGGACACTCACTCCGT Mfn2 CTTTGCTGGGGTGCAAAAGG GAGAGCTCACACAGGGTCAC Opa1 TCTTCACTGCAGGTCCCAAAT CTGACACCTTCCTGTAATGCTTG Drp1 CCAGAGGAACTGGTGTGGTC ACCCCATTCTTCTGCTTCAACT Fis1 GACCCTAGCGTGCTTTCTGT ACACGCAACACTGTAACCCT Gapdh ACCCAGAAGACTGTGGATGG ACATTGGGG GTAGGAACAC

[0070] 1.10 TUNEL staining

[0071] Cell death was quantified using a TUNEL kit provided by Servicebio (China), strictly following the manufacturer's instructions. During sample preparation, the tissue was first treated with proteinase K at 37°C for 22 minutes to enhance tissue permeability. Subsequently, it was treated with 0.1% Triton X-100 at room temperature for 20 minutes to further increase cell membrane permeability. Afterwards, the sample was incubated with a 1:5 diluted TUNEL reagent at 37°C for 2 hours to mark broken DNA. To visualize the cell nucleus, DAPI was used for nuclear staining because it can specifically bind to DNA and emit blue fluorescence under ultraviolet light. All samples were observed under a Leica fluorescence microscope, and the entire process was recorded in real time. Finally, the fluorescent images were quantitatively analyzed using ImageJ software to identify and calculate the extent of cell death.

[0072] 1.11Western blotting (WB)

[0073] C57BL / 6J mice were anesthetized with 2% isoflurane to ensure they remained unconscious throughout the procedure. Following anesthesia, cortical tissue was carefully collected using a brain mold, based on a mouse brain atlas, to minimize damage to surrounding areas. Following tissue collection, proteins were extracted from the cortex using a KeyGEN Biotech kit (China). The extracted protein samples were then denatured, and 30 μg of protein per sample was loaded onto an SDS-PAGE gel for separation and subsequently transferred to a membrane. The membranes were incubated overnight at 4°C with primary antibodies, including: anti-TFAM (from Abcam, UK, 1:1000), rabbit anti-SURF1, NDUFS1, anti-SDHB, anti-UQCRFS1 / RISP, anti-COX5B, anti-ATP5A, anti-PCG-1α, anti-NRF1, anti-Beclin, anti-LC3, anti-PINK1, and anti-Parkin (from Proteintech, USA, 1:1000), rabbit anti-P62 (from Wanleibio, China, 1:1000), and mouse anti-β-actin (from Proteintech, USA, 1:3000). The following day, the membranes were incubated with the appropriate secondary antibodies (goat anti-rabbit or anti-mouse IgG from ZSJQ, China, 1:10,000) for 1 hour at room temperature. The results were then normalized to β-actin levels to assess target protein expression. Protein bands were visualized using a chemiluminescence detection system, and signal intensity was quantified using ImageJ software (NIH, USA).

[0074] 1.12 Flow cytometric analysis

[0075] After extracting and processing cells from mouse cortical tissue, the cell suspension was stained with JC-1 solution (C2006-1, Beyotime, China) and incubated at 37°C for 30 minutes to 1 hour. JC-1 dye displays different fluorescence depending on the mitochondrial membrane potential: in mitochondria with high mitochondrial membrane potential, JC-1 forms aggregates and emits red fluorescence, while in mitochondria with low membrane potential, it remains in monomeric form and emits green fluorescence. After incubation, the cells were washed with PBS to remove unbound dye and resuspended. A flow cytometer was used to excite JC-1 at 488 nm, and fluorescence emission was detected in two channels: FL1 (Em = 525 ± 20 nm) for green fluorescence (low membrane potential) and FL2 (Em = 585 ± 20 nm) for red fluorescence (high membrane potential). Data were analyzed using FlowJo software, and mitochondrial membrane potential was assessed based on fluorescence intensity.

[0076] 1.13 Immunofluorescence staining

[0077] Brain tissue was fixed in 4% formaldehyde solution for 24 hours and then cut into 30 μm thick sections. After fixation, the sections were incubated with 10% goat serum albumin solution for 1 hour to block nonspecific antibody binding. After blocking, the sections were incubated overnight at 4°C with primary antibody solution, including anti-CAD65 (1:200; Abcam, UK), anti-SURF1, LC3, anti-TOMM20 and anti-NeuN (1:200; Proteintech, USA) to label the target proteins. After incubation, the sections were washed three times with 1× PBS, each wash lasting 5 minutes. Subsequently, the sections were incubated with secondary antibodies for 2 hours at room temperature (Cy3-conjugated goat anti-rabbit IgG and FITC-conjugated goat anti-mouse IgG, 1:200; Proteintech, USA). To visualize cell nuclei, DAPI staining was performed. After staining, images were captured using a Leica fluorescence microscope and analyzed using ImageJ software. Fluorescence images were converted into binary images, and threshold processing was applied to identify positively stained areas. Regions of interest (ROIs) within the cortical area were defined based on anatomical landmarks, and quantitative analysis was performed by counting the number of double-labeled cells per square millimeter.

[0078] 1.14 ATP determination

[0079] ATP measurements were performed using an ATP assay kit (S0026, Beyotime) according to the manufacturer's instructions. First, tissue samples were homogenized in lysis buffer to prepare lysates, and then the protein concentration was measured using the BCA method. Next, 100 μl of protein sample was mixed with an equal volume (100 μl) of ATP assay buffer. The mixture was immediately used for luminescence detection. The luminescence signal of the sample was recorded using a luminometer. Based on the intensity of the signal, the ATP concentration was quantified by comparison with a standard curve and expressed as nanomoles of ATP per milligram of protein (nmol ATP / mg protein).

[0080] 1.15DHE staining

[0081] This study used dihydroethidium (DHE) staining to assess reactive oxygen species (ROS) levels. Brain tissue samples were first frozen and cut into 30-μm-thick sections. The sections were then incubated with 5 μM DHE fluorescent dye for 1 hour at 37°C in the dark. After incubation, the sections were washed three times with phosphate-buffered saline (PBS) to remove any unbound dye and then mounted onto microscope slides for sealing. Fluorescence images were captured using a Leica fluorescence microscope, and the images were quantitatively analyzed using ImageJ software to measure the accumulation of ROS.

[0082] 1.16 Statistical Analysis

[0083] All statistical analyses in this study were performed using GraphPad Prism version 9.0 (GraphPad Software, La Jolla, CA, USA). Data are presented as mean ± standard error of the mean (SEM) and are based on the results of at least three independent experiments. For comparisons between two groups, Student's-test was performed to assess statistical significance. For comparisons involving multiple groups, a one-way analysis of variance (ANOVA) was performed first; if significant differences were detected, Tukey's post hoc test was used to determine which specific groups differed. A p-value < 0.05 was considered statistically significant.

[0084] 2. Research Results

[0085] 2.1 Identification of hub genes

[0086] The identification results of differentially expressed genes in the training set are shown in the attached figure. Figure 1 A total of 2,483 differentially expressed genes (DEGs) were identified in the training set, of which 1,510 genes were upregulated and 973 genes were downregulated ( Figure 1 Intersection analysis of DEGs, MD-RGs, and SD-RGs yielded 43 intersection genes that may play a role in MDD, particularly in mitochondrial function and sleep disorders (SD). Figure 1 Middle B). A PPI network of 43 intersection genes was constructed using STRING, containing 43 nodes and 144 edges (confidence > 0.4) ( Figure 1 The top 20 genes were finally identified by evaluating them using the cytoHubba plugin in Cytoscape using six different algorithms (EPC, MCC, Degree, MNC, Closeness, and Radiality). A total of 18 genes were selected by the intersection of the results of these algorithms, including SDHB, MDH2, SOD2, SUCLA2, NDUFS1, GLUD2, DLD, ALDH2, TFAM, HADHB, CYCS, PCCB, SHMT2, CASP3, AFG3L2, LRPPRC, SURF1, and NDUFB9 ( Figure 1 Middle D).

[0087] Then the 18 candidate genes were identified as hub genes, and the results are shown in the attached figure. Figure 2 As shown. First, LASSO regression analysis was performed and the model with the lowest error was selected (lambda.min = 0.007). This process identified 10 characteristic genes: SDHB, SOD2, TFAM, HADHB, CYCS, CASP3, AFG3L2, LRPPRC, SURF1 and NDUFB9 ( Figure 2Subsequently, SVM-RFE analysis was performed on these 18 candidate genes. After recursive feature elimination, the model identified 8 key genes: SURF1, TFAM, SDHB, NDUFB9, NDUFS1, SOD2, AFG3L2, and LRPPRC ( Figure 2 Boruta analysis was used to rank the importance of 18 candidate genes and select characteristic genes, resulting in 13 characteristic genes: SDHB, MDH2, SOD2, NDUFS1, ALDH2, TFAM, HADHB, CYCS, PCCB, CASP3, AFG3L2, SURF1, NDUFB9 ( Figure 2 Middle C). The intersection of genes identified by LASSO, SVM-RFE, and Boruta methods revealed six key genes: SDHB, SOD2, TFAM, AFG3L2, SURF1, and NDUFB9 ( Figure 2 Middle D).

[0088] The expression of six key genes in the transcriptome training set and transcriptome validation set was visualized, and genes with consistent expression trends and significant inter-group differences (p < 0.05) were selected as biomarkers. Finally, three biomarkers were identified: TFAM, SURF1, and NDUFB9 ( Figure 2 In the ROC curve analysis of these two datasets, the AUC values ​​of these three genes were all greater than 0.7, indicating high prediction accuracy ( Figure 2 Therefore, TFAM, SURF1, and NDUFB9 were identified as key genes in this study.

[0089] 2.2 Identification of key cell types and hub gene expression profiles in single-cell transcriptomes

[0090] Based on the UMAP clustering method and marker genes provided in existing literature, this study annotated seven cell types, including astrocytes (Astros), endothelial cells (Endo), excitatory neurons (Ex), inhibitory neurons (Inhib), microglia / macrophages (Micro / Macro), oligodendrocytes (Oligos), and oligodendrocyte precursor cells (OPCs) ( Figure 3 The biomarkers obtained from the transcriptome data were used to observe the differences in distribution between groups and cells in the single-cell data. The results showed that the expression levels of the three biomarkers were higher in the control group compared with the case group, which was consistent with the results of the transcriptome data ( Figure 3Middle D). TFAM is mainly expressed in Oligos, OPCs, Inhib and Ex cells; SURF1 is mainly expressed in Inhib, Ex and Endo cells; NDUFB9 is mainly expressed in OPCs, Inhib and Endo cells ( Figure 3 Intercellular communication was assessed using the CellChatDB database. In both the control and case groups, Ex cells showed stronger interactions with other cells, and the strongest interaction was between Ex and Inhib cells ( Figure 3 Middle F).

[0091] 2.3 Overexpression of SURF1 alleviates CUMS-induced depressive-like behavior in mice

[0092] In this study, multiple experimental methods were used to evaluate the relationship between CUMS-induced depressive-like behavior and key genes. The mRNA expression of TFAM, SURF1, and NDUFB9 was significantly decreased in the CUMS group ( Figure 4 A, B, C, p < 0.01), indicating that the transcriptional activity of these genes was inhibited under chronic stress conditions. Next, we focused on the SURF1 gene, which is expressed in both inhibitory and excitatory neurons. Western Blot analysis further confirmed that the level of SURF1 was significantly reduced in the CUMS group ( Figure 4 D, E, p < 0.01), especially in inhibitory neurons ( Figure 4 (F, G, p < 0.01). These findings indicate that CUMS not only inhibits the transcription and translation of SURF1 but also affects its distribution in neurons.

[0093] Sucrose preference test (SPT) showed that CUMS exposure significantly reduced sucrose preference ( Figure 4 However, SURF1 overexpression reversed this effect and significantly restored sucrose preference (p < 0.01), indicating that SURF1 overexpression has an antidepressant-like effect.

[0094] The forced swim test (FST) showed that the immobility time of the CUMS group was significantly increased ( Figure 4 Overexpression of SURF1 reversed this effect and significantly reduced the immobility time (p<0.01), indicating that overexpression of SURF1 alleviated stress-induced depressive-like behavior.

[0095] These findings were further supported by the open field test (OFT), where CUMS-treated mice showed a significant decrease in total walking distance ( Figure 4 J, p < 0.01), and the number of entries into the central area decreased ( Figure 4K, p < 0.01), both of which are characterized by reduced activity and social avoidance, core features of depression. Overexpression of SURF1 restored these behaviors to baseline levels (p < 0.01), highlighting its potential to alleviate stress-induced behavioral deficits.

[0096] Finally, TUNEL staining showed that overexpression of SURF1 significantly reduced neuronal apoptosis in CUMS-treated brains, suggesting that SURF1 may protect neurons from stress-induced damage ( Figure 4 L and M, p < 0.01).

[0097] These results together highlight the key role of SURF1 in CUMS-induced depressive-like behavior. Overexpression of SURF1 effectively reversed the molecular and behavioral changes induced by CUMS, demonstrating a potential antidepressant effect.

[0098] 2.4 SURF1 overexpression rescues mitochondrial abnormalities in the PFC of CUMS mice. This study comprehensively evaluated the effects of SURF1 overexpression on CUMS-induced depressive-like behaviors. Furthermore, this study also paid special attention to its effect on the restoration of mitochondrial function.

[0099] First, regarding the expression of mitochondrial electron transport chain complexes, the results showed that the expression of these complexes was significantly reduced in CUMS-exposed mice. Specifically, the expression levels of complexes I (NDUFS1), II (SDHB), III (UQCRFS1 / RISP), IV (COX5B), and V (ATP5A) were significantly decreased, indicating stress-induced mitochondrial dysfunction. However, in mice with SURF1 overexpression, the expression levels of these complexes were significantly restored to near normal levels, with statistical significance ( Figure 5 AG, p < 0.01), demonstrating that SURF1 overexpression can effectively restore the function of the electron transport chain.

[0100] Further evaluation of mitochondrial function also confirmed the protective effect of SURF1 overexpression. ATP was measured in different groups of mice. In CUMS-exposed mice, ATP levels in the prefrontal cortex (PFC) were significantly reduced, reflecting impaired mitochondrial energy metabolism. In contrast, mice overexpressing SURF1 successfully restored ATP levels close to normal ( Figure 5 In addition, measurements of mitochondrial membrane potential showed a significant decrease in CUMS mice, indicating mitochondrial dysfunction and potential energy depletion ( Figure 5I and J, p < 0.01). However, mice overexpressing SURF1 maintained a relatively high membrane potential, indicating that SURF1 effectively maintains the integrity and function of the mitochondrial membrane. Regarding oxidative stress, ROS levels were significantly elevated in CUMS mice, pointing to mitochondrial dysfunction and enhanced cellular stress responses ( Figure 5 In contrast, ROS levels were significantly reduced in mice overexpressing SURF1, indicating that SURF1 overexpression can alleviate oxidative stress and protect neurons from damage.

[0101] Taken together, these results suggest that SURF1 overexpression not only restores the expression of mitochondrial electron transport chain complexes but also enhances ATP production, maintains mitochondrial membrane potential, and reduces oxidative stress. These effects collectively alleviate CUMS-induced depressive-like behavior, further supporting the protective role of SURF1 in depression models.

[0102] 2.5SURF1 overexpression restores mitochondrial biogenesis and mitophagy in the PFC of CUMS mice

[0103] In this study, we also investigated the protective mechanism of SURF1 on mitochondrial function in mice under CUMS conditions. The PGC-1α / nuclear respiratory factor 1 (NRF1) / mitochondrial transcription factor A (TFAM) signaling pathway plays a key role in the regulation of mitochondrial biogenesis. Further analysis showed that CUMS significantly reduced the expression of PGC-1α, NRF1, and TFAM ( Figure 6 AD, p < 0.01). This suggests that stress may impair mitochondrial function by inhibiting mitochondrial biogenesis. However, overexpression of SURF1 effectively restored the expression of these proteins, indicating that SURF1 can protect mitochondrial function by promoting mitochondrial biogenesis.

[0104] In terms of mitochondrial autophagy, this study measured the levels of autophagy-related proteins P62, LC3, and Beclin. The results showed that CUMS treatment significantly increased the expression of P62, while the LC3II / LC3I ratio decreased and Beclin expression decreased, indicating that autophagy was inhibited ( Figure 6 (JM, p < 0.01). Notably, overexpression of SURF1 significantly restored the expression levels of these proteins, indicating that SURF1 enhances autophagy and promotes the clearance of damaged mitochondria, thereby improving mitochondrial quality control.

[0105] Immunofluorescence analysis further confirmed these findings. The colocalization of LC3 (an autophagy marker) and TOMM20 (a mitochondrial marker) was significantly reduced in the cortical tissue of CUMS mice compared with controls ( Figure 6However, overexpression of SURF1 significantly increased the colocalization of these two markers (P < 0.01), further confirming that SURF1 promotes autophagy.

[0106] In addition, this study also evaluated the PINK1 / Parkin-mediated autophagy pathway. We found that CUMS treatment significantly reduced the protein levels of PINK1 and Parkin ( Figure 6 However, overexpression of SURF1 reversed this decrease and restored the expression of PINK1 and Parkin, indicating that SURF1 regulates autophagy through the PINK1 / Parkin pathway.

[0107] 2.63-MA and Mdivi-1 inhibited the protective effect of SURF1 against mitochondrial dysfunction induced by CUMS

[0108] In this study, we used the mitophagy inhibitors 3-MA and Mdivi-1 to intervene in mice and further evaluated the role of SURF1 in alleviating the depressive-like behavior caused by CUMS. Immunofluorescence analysis showed that 3-MA and Mdivi-1 significantly reduced the colocalization of LC3 and TOMM20 ( Figure 7 A, B, p < 0.01), indicating that these inhibitors effectively inhibited the mitochondrial autophagy process. Further Western blot analysis showed that the treatment of these inhibitors significantly reduced the protein expression levels of mitochondrial electron transport chain complex I (NDUFS1), II (SDHB), III (UQCRFS1 / RISP), IV (COX5B) and ATP synthase (ATP5A) in the cortical tissue of mice ( Figure 7 medium CI, p < 0.01), indicating that mitochondrial function was impaired and the activity of electron transport chain complexes was reduced.

[0109] In the assessment of mitochondrial function, ATP measurement results showed that 3-MA or Mdivi-1 treatment effectively prevented the increase in ATP levels induced by SURF1 overexpression ( Figure 7 J, p<0.01), indicating that inhibition of mitophagy had a significant effect on ATP synthesis. Similarly, mitochondrial membrane potential analysis showed that the inhibitor treatment group failed to restore the normal mitochondrial membrane potential caused by SURF1 overexpression ( Figure 7 K, p<0.01), indicating that the ability to maintain membrane potential was severely disturbed. In addition, ROS measurement showed that 3-MA or Mdivi-1 treatment inhibited the decrease in ROS levels induced by SURF1 overexpression ( Figure 7This finding further suggests that the regulation of oxidative stress is affected, supporting the role of mitophagy in the antidepressant effects of SURF1. Treatment with the inhibitor significantly disrupted SURF1's ability to regulate mitochondrial function through mitophagy, providing strong evidence that mitophagy plays a key role in the antidepressant effects of SURF1.

[0110] Taken together, this study provides evidence for a novel role for SURF1 in the pathophysiology of major depressive disorder (MDD). Our findings suggest that SURF1 may promote mitochondrial biogenesis and mitophagy, thereby restoring mitochondrial function and ameliorating depressive-like behaviors. These findings contribute to a deeper understanding of the pathophysiology of MDD and provide potential therapeutic targets for the treatment of MDD associated with sleep disorders (SD).

[0111] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. Application of mitochondrial differentially expressed genes in the preparation of diagnostic agents and therapeutic drugs for depression, characterized in that: The mitochondrial differentially expressed genes include one or more of TFAM, SURF1 and NDUFB9.

2. The use according to claim 1, characterized in that: The mitochondrial differentially expressed gene is SURF1.

3. The use according to claim 2, characterized in that: SURF1 mRNA expression is reduced in depression.

4. The research method of application according to any one of claims 1 to 3, characterized in that: The following steps are included: S1: Identification of differentially expressed mitochondrial genes associated with depression by transcriptomics and single-cell transcriptomics sequencing technologies; S2: The relationship between SURF1 overexpression and CUMS-induced depressive-like behavior in mice was verified by sucrose preference test, forced swimming test and open field test; S3: Analysis of mitochondrial function in mice by SURF1 overexpression and restoration of CUMS-induced mitochondrial function; S4: The protective effect of SURF1 against mitochondrial dysfunction caused by CUMS was further verified by targeted inhibitors.

5. Use of reagents for detecting SURF1 gene in the preparation of diagnostic agents for depression.

6. Use of SURF1 gene activator in the preparation of drugs for treating depression.

7. Use of a product containing SURF1 protein in the preparation of a drug for treating depression.

Images