Single-nucleus profiling of the left ventricle of the mouse heart after chronic stress

Abstract

It is widely accepted that stress and cardiac disease are related, but the exact mechanism is still up for debate. Although stress has been scientifically confirmed as a causative factor, there is still a lack of qualitative and quantitative indicator systems for stress-induced cardiac injury. The forensic evidence frequently indicates that cases of sudden cardiac death are often preceded by prolonged exposure to chronic stress factors. However, the cellular responses and mechanisms that trigger and regulate these activities in the pathological and physiological processes of chronic stress are still poorly understood. The left ventricle (LV), as a critical target organ in cardiovascular diseases, still has unclear cellular heterogeneity under chronic stress. Using single-nucleus RNA sequencing (snRNA-seq), we established a cellular atlas of 99,154 LV cells (42,491 stress and 56,663 control), comprehensively profiling all major cardiac cell types. The resulting dataset not only provides a foundation for deciphering the molecular mechanisms underlying chronic stress in cardiovascular dysfunction but also enables the identification of potential biomarkers for future diagnostic and therapeutic strategies.

Background & Summary

In present society, people inevitably face various stressful situations, such as work pressure, economic burden, interpersonal tension, and health problems. Acute stress can lead to arrhythmia, myocardial infarction, Takotsubo cardiomyopathy and even sudden death, and chronic stress has been shown to be associated with diseases such as atherosclerosis and hypertension¹. The term stress describes the series of fight or flight responses stimulated by various stimulus (stressors). The activation of the locus ceruleus-norepinephrine/sympathetic-adrenal medulla (LC/NE) and hypothalamus-pituitary-adrenal (HPA) axis triggers the systemic biological response to stress, increasing levels of catecholamines (CA) like dopamine, adrenaline, and norepinephrine, as well as glucocorticoids like cortisol. These responses guarantee the proper function and metabolism of vital organs and have a positive impact on defense and compensation². Despite diagnosed with myocardial ischemia or even sudden cardiac death by the clinical diagnosis or autopsy determination, a significant proportion of patients (decedents) are not found to have severe cardiovascular pathological changes^3,4. Although some studies have confirmed that stress-induced acute catecholamine surge, vascular inflammation, and endothelial activation may underlie cardiovascular pathologies such as stress cardiomyopathy and atherosclerosis^5,6, the potential stress-induced changes have not been fully investigated. Because there is no well-defined mechanism or set of markers for chronic stress-related cardiovascular injury, the role of chronic stress in the diagnosis, treatment, and cause of death of diseases remains controversial.

The LV, as the main compensatory chamber of the heart, is a major target of common cardiovascular diseases⁷. However, the research on LV cell heterogeneity under chronic stress is still blank. Current research mostly focuses on overall cardiac function or specific cells (such as cardiomyocytes), but chronic stress may lead to early phenotypic changes in non-cardiomyocytes (such as endothelial cells and immune cells), which have not been systematically characterized⁸. And most single-cell studies have small sample sizes (<50000 cells), making it difficult to detect rare subgroups⁹. Our data fills the gap in the dynamic response of LV whole cell types under chronic stress; The single-cell resolution of nearly 100000 cells is sufficient to analyze the role of various cell subpopulations in chronic stress.

This dataset provides comprehensive transcriptional profiles of LV cell populations derived from both chronically stressed and control mice, captured at a resolution sufficient to facilitate detailed analysis of cellular subpopulations. These data are amenable to applications such as LV cell clustering, identification of novel cellular markers, and investigation of molecular mechanisms underlying chronic stress-induced cardiac pathology, making it a valuable resource for researchers in this field.

Methods

Animals and treatments

Eight-week-old male C57BL/6 N mice (body weight 20–22 g) were obtained from the Experimental Animal Center of Peking University Health Science Center (Beijing, China) and housed in the Animal Experiment Center of the Forensic Medicine Department, Hebei Medical University. The animals were kept in a controlled environment with a temperature of 23 ± 2 °C, a 12 h/12 h light-dark cycle, and 50% humidity. Water and regular rodent feed were available to all animals without charge. Prior to experimentation, the animals were acclimatized to the new environment for 7 days. The Institutional Animal Care and Use Committee of Hebei Medical University (Approval No. 20223011) approved the Guidelines for the Care and Use of Laboratory Animals, and all experimental procedures were carried out in compliance with the applicable rules and regulations. The animals were split into two groups at random: the chronic stress 28-day group (CS, n = 4) and the control group (CON, n = 4). Mice in the CS group were subjected to chronic unpredictable mild stress (CUMS)¹⁰ for 28 days according to the protocol established by Hui Ma et al.¹¹, while CON group mice were maintained under standard feeding conditions. On the day following the completion of chronic stress protocol, all animals were subjected to open field test (OFT) after 1-hour habituation in a dark room to evaluate the successful establishment of the stress model¹².

Validation of the CUMS model

The successful establishment of the CUMS model was validated through behavioral assessment using the OFT¹³. Specifically, To assess the effectiveness of stress induction, we measured the total distance traveled and the percentage of time spent in central, which are trustworthy markers of exploratory behavior and anxiety-like reactions. Compared with the CON group, The reduction in total distance traveled and the decrease in the percentage of residence time in central observed in CS-group mice collectively confirmed successful establishment of the stress model.

Nuclear isolation procedure

Samples of frozen heart tissue were thawed on ice, cut into tiny pieces, and then washed with ice-cold PBS. To pellet the cells, the cell solution was centrifuged at 50 g for two minutes at 4 °C. The cells were then resuspended in two milliliters of lysis buffer. Cell homogenization was performed using a bead homogenizer at 5,000 rpm for 20 s. Next, 10 μL of 4.5 mg/mL digitonin was added to 1 mL of cell suspension and incubated at 4 °C for 30 min. The semi-permeabilized suspension was transferred to a 7 mL Dounce homogenizer, and tissue disruption was carried out with 30 strokes using the loose pestle, followed by 30 strokes with the tight pestle. Samples were sonicated at 100% amplitude (130 W) with 5 s pulses and 5 s intervals for a total duration of 20 s. The homogenate was centrifuged at 100 g for 5 min at 4 °C to remove unbroken cells, tissue debris, and large cellular fragments. The supernatant was sequentially filtered through 40 μm and 20 μm cell strainers, and the filtrate was centrifuged at 2,000 g for 10 min at 4 °C to pellet nuclei. After each wash, the nuclear pellet was centrifuged at 2,000 g for 10 min at 4 °C using 5 mL of nuclear storage buffer. Prior to further processing, purified nuclei were kept on ice for up to 12 hours.

Quality control and data preprocessing

Quality control and basic statistical analysis of raw sequencing reads were performed using fastp. The FASTQ-formatted raw reads generated from the Illumina platform were preprocessed using Trimmomatic software. Low-quality reads were removed by implementing a 4-base sliding window approach, with reads being truncated when the average base quality within the window fell below 10 (SLIDINGWINDOW:4:10). Trailing low-quality bases and N bases were trimmed from read ends, removing regions with quality scores below 3 (TRAILING:3). Adapter sequences were eliminated through a dual-strategy approach: first by requiring alignment with adapter sequences showing at least 7 matching bases with no more than 2 mismatches, and second by removing non-overlapping regions when the overlapping base quality score between paired reads exceeded 30 (ILLUMINACLIP:adapter.fa:2:30:7). Subsequent filtering removed reads shorter than 26 bases and those failing to form proper read pairs. Reads that successfully passed all quality filtering steps were designated as clean reads and used for downstream analyses. Final quality assessment statistics for the processed clean reads were generated using fastp.

10× Genomics snRNA-seq analysis

Novogene (Beijing, China) prepared the 10 × —Genomics library and carried out the snRNA-Seq. After two PBS washes and a 30-minute room temperature incubation period, nuclei were loaded onto 3′ v3 chromium microfluidic chips. utilizing the 10 × — Chromium Controller (10X Genomics) for barcoding. Subsequently, barcoded RNA was reverse-transcribed, and sequencing libraries were constructed using reagents from the Chromium Single Cell 3′ v3 Kit (10X Genomics) according to the manufacturer’s protocol. Sequencing was performed on the Illumina NovaSeq 6000 platform following the manufacturer’s instructions (Illumina). Raw reads were demultiplexed and mapped to the mouse reference genome (GRCm38, version M23, Ensembl: version 92) using the 10X Genomics Cell Ranger pipeline (https://support.10xgenomics.com/single-cell-gene-expression/software/pipelines/latest/what-is-cell-ranger). All downstream single-cell analyses were performed using Cell Ranger and Seurat¹⁴. Briefly, unique molecular identifiers (UMIs) were counted for each gene and each cell barcode (filtered by CellRanger) to construct a digital expression matrix. Secondary filtering in Seurat was applied: genes expressed in more than three cells were considered expressed, and each cell was required to have at least 200 expressed genes. Potential doublets and low-quality cells were filtered out. Data normalization, dimensionality reduction, clustering, and differential expression analysis were performed using the Seurat package. For integrated analysis of datasets, we employed the canonical correlation analysis (CCA) alignment method in Seurat¹⁵. Highly variable genes were chosen for clustering, and a graph-based clustering approach with a resolution parameter of 0.6 was applied utilizing principal components obtained from these genes.

Inter-sample global analysis

To identify region-specific marker genes, inter-sample differential expression analysis was carried out using the edgeR package¹⁶ based on the gene expression matrix that Seurat had filtered.

Marker gene enrichment analysis

Gene Ontology (GO) enrichment analysis of marker genes was performed using the clusterProfiler R package, with correction for gene length bias¹⁷. GO terms with adjusted p-values less than 0.05 were considered significantly enriched by the marker genes. KEGG¹⁸ is a database resource for understanding high-level functions and utilities of biological systems, such as cells, organisms, and ecosystems, from molecular-level information, particularly large-scale molecular datasets generated by genome sequencing and other high-throughput experimental technologies (http://www.genome.jp/kegg/). We tested the statistical enrichment of marker genes in KEGG pathways using the clusterProfiler R package. Marker gene analysis based on Reactome pathways was conducted using the ReactomePA R package. REACTOME is an open-source, open-access, manually curated, and peer-reviewed pathway database (https://reactome.org/).

Data Records

The raw data generated in this study have been deposited in the Gene Expression Omnibus (GEO) database under the accession number GSE263648¹⁹. For the control group, the sample IDs are GSM8195986–GSM8195989; the chronic stress group IDs are GSM8195990-GSM8195993. The compressed gene expression matrix file ‘filtered_feature-bc_matrix. h5’ has been stored in Figshare, which contains cell barcodes, genes and corresponding expression values (https://doi.org/10.6084/m9.figshare.28784090.v2)²⁰. The analysis results data, including differential expression and cell type annotation have been deposited in Figshare (https://doi.org/10.6084/m9.figshare.28847951.v1)²¹.

Technical Validation

We constructed snRNA-seq libraries using the 10× Genomics method and performed sequencing on the Illumina NovaSeq 6000 platform (Fig. 1a). The total reads for both the CON group and the CS group exceeded 355 million, with valid barcode detection rates surpassing 96% (Table 1). The detailed QC and FASTQ files for each sample are shown in Table 2. The median number of genes detected per cell was 1,496, with an average of 34,752 reads per cell. The batch effects between samples were minimized using the Seurat algorithm. The sequencing quality of both groups was high and comparable, indicating minimal technical bias introduced in this study. Overall, we obtained high-quality single-nucleus transcriptomic datasets of the mouse LV from both the CS and CON groups. The estimated numbers of detected cells in the CS and CON groups were 42,491 and 56,663, respectively, with average gene counts per cell of 1,202 and 1,789, respectively. Other statistical metrics of the single-cell level sequencing data are presented in Table 1 and Fig. 2. These measurements showed a decent proportionality of our sequencing data, with the two groups being equal. Therefore, it is appropriate to use these sequencing data to conduct the transcriptome analysis of both groups. Additionally, we provided sequencing statistics for each sample to confirm that the quality control of each sample met the standards (Table 2). In addition, in order to obtain more comprehensive gene expression information, we followed the official guidelines for 10xGenomics single-cell 3′V3 and made the sequencing depth of each cell exceed 20Kreads according to the library requirements. Therefore, our sequencing depth is sufficient for subsequent analysis.

[See PDF for image]

Fig. 1

Quality control (QC) analysis of mouse left ventricle snRNA-seq data. (a) Schematic workflow of single-nucleus RNA sequencing. (Created with Biorender.com). (b) Scatter plots illustrating the number of detected genes per cell (left), unique molecular identifiers (UMIs) per cell (middle), and the percentage of mitochondrial gene expression (right) across both samples. (c) UMAP visualization showing the distribution of mRNA counts (left panel) and gene counts (right panel) across the single-cell population.

Table 1. Sequencing and Cell Ranger statistics.

Sample_Name	CON	CS
Estimated.Number.of.Cells	56,663	42,491
Mean.Reads.per.Cell	25,211	44,292
Median.Genes.per Cell	1,202	1,789
Number.of.Reads	355,640,333	355,853,590
Valid.Barcodes	96.25%	96.33%
Sequencing.Saturation	66.48%	65.93%
Q30.Bases.in.Barcode	96.08%	95.95%
Q30.Bases.in.RNA.Read	88.10%	88.88%
Q30.Bases.in.UMI	93.83%	93.60%
Reads.Mapped.to.Genome	75.28%	83.33%
Reads.Mapped Confidently.to.Genome	73.63%	81.65%
Reads.Mapped.Confidently.to.Intergenic.Regions	5.63%	5.90%
Reads.Mapped.Confidently.to.Intronic.Regions	47.10%	51.45%
Reads.Mapped.Confidently.to.Exonic.Regions	20.90%	24.30%
Reads.Mapped.Confidently.to.Transcriptome	51.55%	57.40%
Reads.Mapped.Antisense.to.Gene	16.18%	18.03%
Fraction.Reads.in.Cells	64.85%	68.80%
Total Genes.Detected	26,375	25,944
Median.UMI.Counts.per.Cell	1,840	3,376

Table 2. Detailed QC of FASTQ files.

LibraryID	Sample Name	Raw Reads	Raw Bases (G)	Q20 (%)	Q30 (%)	GC Content (%)
FRHX23H003167-1a	CON1	355367366	106.61	94.56	89.06	44.45
FRHX23H003168-1a	CON2	344102318	103.23	93.01	86.91	45.19
FRHX23H003169-1a	CON3	365512686	109.65	93.79	88.12	44.74
FRHX23H003170-1a	CON4	357578960	107.27	94.07	88.27	44.37
FRHX23H003163-1a	CS1	401540711	120.46	95.03	89.37	42.4
FRHX23H003164-1a	CS2	350516308	105.15	94.24	88.23	43.01
FRHX23H003165-1a	CS3	335409613	100.62	94.99	89.47	45.1
FRHX23H003166-1a	CS4	328034311	98.41	94.12	88.4	44.39

[See PDF for image]

Fig. 2

Sequencing saturation and median genes per Cell. (a) Curves showing the sequencing saturation (left) and the median number of genes per cell (right) of each sample in the CON group. (b) Curves showing the sequencing saturation (left) and the median number of genes per cell (right) of each sample in the CS group.

After obtaining the gene expression matrix, we filtered the cells again based on indicators such as the number of detected genes and the proportion of mitochondrial UMI. We also used DoubletFinder to predict the doublets in snRNA sequencing data, removing multiple cells from the dataset to ensure the reliability and accuracy of subsequent analysis results. A threshold percentage has been established. Necrotic and apoptotic cells should be roughly eliminated from the data if mito <0.1. Furthermore, cells with fewer than 200 genes identified are classified as necrotic or apoptotic, while cells with more than 10000 genes are regarded as binary; both of these categories have been eliminated from our data (Fig. 1b,c). Table 3 displays the number of genes found in each sample cell together with CellQC data. The majority of the cells were kept after all of these screening, proving the high caliber of our sample preparation and sequencing procedure and the accuracy of the data.

Table 3. The number of genes detected in each sample cell and Cell QC.

Sample_Name	CON1	CON2	CON3	CON4	CS1	CS2	CS3	CS4
low.thresholds_nGene	200	200	200	200	200	200	200	200
high.thresholds_nGene	4000	3500	3500	4000	5000	6500	5000	3500
high.thresholds_percent.HB	5	5	5	5	5	5	5	5
high.thresholds_percent.mito	5	10	10	10	10	10	10	10
min.cells	3	3	3	3	3	3	3	3
gene_number	25945	26368	26531	26656	25746	25598	26388	26023
gene_number_filtered	22788	23111	23344	23483	22823	22301	23199	22921
nGene_median	1339	1036	1160	1270	1683	2533	1695	1233
nGene_median_filtered	1283	994	1110	1218	1600	2504	1649	1168
cell_number	12676	14197	14627	15163	14940	3911	8825	14815
cell_number_filtered	11418	12759	13086	13473	13241	3754	8194	13182

To characterize cellular heterogeneity and validate the robustness of our dataset, we performed dimensionality reduction using principal component analysis (PCA) followed by uniform manifold approximation and projection (UMAP) (Fig. 3a). Comparative analysis of cellular distributions between the two experimental groups demonstrated high reproducibility across sequencing replicates (Fig. 3b). 23 different cell clusters were found by unsupervised clustering analysis (Fig. 3a, left panel). Cell type identification and differential gene screening were performed based on top differentially expressed genes (DEGs), with quantitative criteria set as P < 0.05, |log2FC| > 0.25. Additionally, genes were required to exhibit an expression rate >10% in the target cell population. Ultimately, we identified and defined seven distinct cell populations: endothelial cells (Pecam1+, Vwf+, Dcn+), ventricular cardiomyocytes (Myh7+, Myl2+, Fhl2+), fibroblasts (Dcn+, Gsn+, Pdgfra+), pericytes (Rgs5+, Abcc9+, Kcnj8+), mesothelial cells (Msln+, Wt1+, Bnc1+), myeloid cells (C1qa+), and neurons (Plp1+) (Fig. 3a, right panel). Figure 3c,d show the established expression profiles of marker genes in the aforementioned cardiac cell population. The high expression of each identified marker gene in specific cell types further demonstrates the reliability of cell type identification.

[See PDF for image]

Fig. 3

snRNA sequencing reveals cellular heterogeneity in the murine left ventricle. (a) UMAP visualization of unsupervised clustering analysis identifying 23 distinct cell populations. Cell type annotation was performed based on established cell-type-specific marker genes. (b) Integrated UMAP projection demonstrating the spatial distribution of left ventricular cells across eight independent samples. (c) Hierarchically clustered heatmap displaying the top differentially expressed marker genes (adjusted p-value < 0.05) for the seven major cell populations. (d) Violin plots illustrating the expression patterns of canonical marker genes across annotated cell types. (e) Violin plots depicting the expression profiles of known marker genes within each identified cluster.

In summary, we have a clean gene-cell expression matrix with clustering information based on stringent quality control and data filtering. The resulting dataset is ideal for downstream analytics, CS cell biology research, and the identification of new therapeutic and diagnostic targets.

Usage Notes

The dataset can be used independently (but not limited to) (1) to analyze gene expression profiles of each cell type in the LV of mice under CS and identifying potential diagnostic/therapeutic targets, (2) to investigate cell-cell interactions within the LV following CS. (3) to construct a gene regulatory map within left ventricular cells following CS by analyzing the interactions between transcription factors (TFs) and target genes.

Acknowledgements

This work was supported by the Major Program of National Natural Science Foundation of China (82293650/82293651), the Key Program of National Natural Science Foundation of China (82130055), Science and technology research project of colleges and universities in Hebei Province (QN2016181).

Author contributions

Bin Cong, Qian Qi, Yaping Li, Xia Liu and Qian Wang designed the experimental model. Yaping Li, Han Xiao, Qinmin Chen and Shulin Xiang contributed to the sample preparation/tissue collection. Yaping Li wrote the manuscript with inputs from all authors. Haibo Xu, Qinmin Chen, Yan Zhu and Huaxing Zhang performed data analyses.

Code availability

The code used to process the raw sequencing files and generate all the results pres-ented in this study can be found in https://github.com/Yaping116/Code-for-Single-Cell-Analysis.

Competing interests

The authors declare no competing interests.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Ajibewa, TA et al. Chronic stress and cardiovascular events: findings from the cardia study. Am. J. Prev. Med.; 2024; 67, pp. 24-31. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38143043]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=38143043&query_hl=1

2. Steptoe, A; Kivimaki, M. Stress and cardiovascular disease. Nat. Rev. Cardiol.; 2012; 9, pp. 360-370.1:CAS:528:DC%2BC38Xns1SktLw%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22473079]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=22473079&query_hl=1

3. Lanza, GA. Angina pectoris and myocardial ischemia in the absence of obstructive coronary artery disease: role of diagnostic tests. Curr. Cardiol. Rep.; 2016; 18, 15. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26768741]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=26768741&query_hl=1

4. Rahman, H et al. Coronary microvascular dysfunction is associated with myocardial ischemia and abnormal coronary perfusion during exercise. Circulation; 2019; 140, pp. 1805-1816. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31707835][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6882540]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=31707835&query_hl=1

5. Hinterdobler, J et al. Acute mental stress drives vascular inflammation and promotes plaque destabilization in mouse atherosclerosis. Eur. Heart J.; 2021; 42, pp. 4077-4088.1:CAS:528:DC%2BB3MXisFOgtLjM [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34279021][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8516477]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=34279021&query_hl=1

6. Khalid, N., Shams, P., Shlofmitz, E. & Chhabra, L. Pathophysiology of takotsubo syndrome, http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=30844187&query_hl=1 (2025).

7. Mongirdiene, A. et al. Relationship between oxidative stress and left ventricle markers in patients with chronic heart failure. Cells12, http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=36899939&query_hl=1 (2023).

8. Jiang, Z et al. Dipeptidyl peptidase-4 deficiency prevents chronic stress-induced cardiac remodeling and dysfunction in mice. FASEB. J.; 2025; 39, 1:CAS:528:DC%2BB2MXkt1Ojtr8%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39968759]e70398.http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=39968759&query_hl=1

9. Heimberg, G; Bhatnagar, R; El-Samad, H; Thomson, M. Low dimensionality in gene expression data enables the accurate extraction of transcriptional programs from shallow sequencing. Cell Syst.; 2016; 2, pp. 239-250.1:CAS:528:DC%2BC2sXhtFKksrg%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27135536][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4856162]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=27135536&query_hl=1

10. Du Preez, A et al. Chronic stress followed by social isolation promotes depressive-like behaviour, alters microglial and astrocyte biology and reduces hippocampal neurogenesis in male mice. Brain. Behav. Immun.; 2021; 91, pp. 24-47. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32755644]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=32755644&query_hl=1

11. Ma, H. et al. Amygdala-hippocampal innervation modulates stress-induced depressive-like behaviors through ampa receptors. Proc. Natl. Acad. Sci. USA. 118, http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=33526688&query_hl=1 (2021).

12. Kraeuter, A; Guest, PC; Sarnyai, Z. The open field test for measuring locomotor activity and anxiety-like behavior. Methods Mol Biol; 2019; 1916, pp. 99-103.1:CAS:528:DC%2BC1MXitVyrt7rJ [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30535687]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=30535687&query_hl=1

13. Yoshizaki, K., Asai, M. & Hara, T. High-fat diet enhances working memory in the y-maze test in male c57bl/6j mice with less anxiety in the elevated plus maze test. Nutrients12, http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=32659954&query_hl=1 (2020).

14. Macosko, EZ et al. Highly parallel genome-wide expression profiling of individual cells using nanoliter droplets. Cell; 2015; 161, pp. 1202-1214.1:CAS:528:DC%2BC2MXpt1Sgt7o%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26000488][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4481139]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=26000488&query_hl=1

15. Haghverdi, L; Lun, ATL; Morgan, MD; Marioni, JC. Batch effects in single-cell rna-sequencing data are corrected by matching mutual nearest neighbors. Nat. Biotechnol.; 2018; 36, pp. 421-427.1:CAS:528:DC%2BC1cXmslKrtLo%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29608177][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6152897]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=29608177&query_hl=1

16. Trapnell, C et al. Differential gene and transcript expression analysis of rna-seq experiments with tophat and cufflinks. Nat. Protoc.; 2012; 7, pp. 562-578.1:CAS:528:DC%2BC38Xjt1Cjsrc%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22383036][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3334321]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=22383036&query_hl=1

17. Combes, F; Loux, V; Vandenbrouck, Y. Go enrichment analysis for differential proteomics using proteore. Methods Mol Biol; 2021; 2361, pp. 179-196.1:CAS:528:DC%2BB3sXisVClsLjN [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34236662]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=34236662&query_hl=1

18. Kanehisa, M; Goto, S. Kegg: kyoto encyclopedia of genes and genomes. Nucleic. Acids. Res.; 2000; 28, pp. 27-30.1:CAS:528:DC%2BD3cXhvVGqu74%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10592173][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC102409]http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=10592173&query_hl=1

19. Li, Y., Cong, B. & Qi, Q. Geo. https://identifiers.org/geo/GSE263648.

20. Li, Y. Single-nucleus profiling of the left ventricle of the mouse heart after chronic stress 2025; 10.6084/m9.figshare.28784090.v2figshare;

21. Li, Y. Results 2025; 10.6084/m9.figshare.28847951.v1figshare;