Clinical research ethics statement. The bone marrow tissue used in this study was collected from patients diagnosed with aplastic anemia (AA) who underwent a biopsy at our hospital from January 2020 to January 2022. The study involved 5 patients with AA and 2 healthy volunteers. Before participating in this study, all participants had to sign a written informed consent form. The age range of the patients was from 43 to 65, with an average age of 54. The gathered tissue is divided into two parts. One part is promptly preserved in liquid nitrogen, while the other is fixed in 10% formaldehyde and embedded in paraffin for sectioning [22]. This study has been approved by the Clinical Ethics Review Committee at our hospital (KYLL-2019-023) in accordance with the Helsinki Declaration.
Preparation and sequencing of single-cell samples. We will select bone marrow samples from two healthy adult donors and two patients with AA from the samples collected at our hospital. Bone marrow samples were digested using 1 mg/mL of STEMxyme1 (Worthington, LS004106) and 1 mg/mL of Dispase II (ThermoFisher Scientific, 17105041). Then, they were incubated with 2% fetal bovine serum (FBS, Gibco, 10091148) in culture medium 199 (Gibco, 11150059) at 37°C for 25 minutes. After digestion, the sample was filtered through a 70 µm cell filter (Fisher Scientific, 08-771-2) and collected in a separate tube [23].
Once the samples are prepared as single-cell suspensions, the C1 Single-Cell Auto Prep System (Fluidigm, C1) is employed to capture individual cells. Once the single cells are captured, they undergo lysis within the chip to liberate mRNA. Subsequently, this mRNA is reverse-transcribed to produce complementary DNA (cDNA). The complementary DNA (cDNA) is pre-amplified in a microfluidic chip after lysis and reverse transcription, in preparation for subsequent sequencing. Library construction will be carried out on the amplified cDNA, followed by single-cell sequencing using the HiSeq 4000 Illumina platform. The sequencing parameters will include paired-end reads with a read length of 2×75 bp and approximately 20,000 reads per cell [24].
TSNE clustering analysis, cell annotation and pseudotime analysis. This article analyzes single-cell RNA sequencing (scRNA-seq) data using the Seurat package in the R software. Initially, a series of quality controls were conducted, with the corresponding filtering thresholds set as follows: nFeature_RNA > 200, nFeature_RNA < 5000, and percent.mt < 10. The canonical correlation analysis (CCA) method was employed to eliminate batch effects, followed by normalization of the data using the LogNormalize function. To reduce the dimensionality of the scRNA-Seq dataset, principal component analysis (PCA) was performed on the highly variable genes using the top 2000 genes with the highest variance. Subsequently, the top 30 principal components were selected for TSNE clustering analysis. To identify major cell subpopulations, the FindClusters function provided by Seurat should be employed with the default resolution set at 0.9 (res = 0.9). Next, the t-SNE algorithm is applied to non-linearly reduce the dimensionality of scRNA-seq sequencing data. Filtering marker genes for different cell subpopulations using the Seurat package. To further annotate the marker genes of each cell cluster, we utilized the "SingleR" package and loaded the reference dataset using the HumanPrimaryCellAtlasData function. We annotated the cells by combining well-known marker genes specific to cell lineages with the online resource CellMarker [25].
Subsequently, we conducted pseudotemporal analysis using the "monocle" package in the R software and examined cell communication using the "cellchat" package.
Spatial transcriptome sequencing. Spatial transcriptomic analysis was conducted on the bone marrow of the remaining three patients diagnosed with AA, utilizing the 10x Genomics Visium platform. The 10 µm tissue sections from fresh frozen human bone marrow embedded in OCT were placed onto Visium spatial slides. It was followed by a 30-minute permeabilization step to release mRNA. The mRNA molecules attach to the barcode oligonucleotides positioned at the bottom of the slide and undergo reverse transcription following the instructions provided by the manufacturer. The cDNA libraries prepared using these samples were sequenced on the Illumina NextSeq 2000 platform, generating over 50,000 reads per position. Each library produced more than 400 million reads. We utilized the Spaceranger software (version 3.1.0, 10x Genomics) to align each Visium spatial transcriptomics slide location and acquire raw counts. Subsequently, these counts were compared to the reference data of the GRCh38 human genome.
The 10x Visium spatial transcriptomics data analysis involves using the Load10X_spatial function in Seurat to integrate the raw gene expression matrix, spatial information, and tissue H&E images and create a Seurat object. Before conducting principal component analysis, the data was normalized using SCTransform. The dimensionality was subsequently reduced by selecting the top 30 principal components. To detect marker genes and perform differential gene expression analysis, the FindAllMarkers function in Seurat is used. To identify genes that exhibit spatial variation in situ, the FindSpatiallyVariableFeatures function was employed with default settings.
We implemented spatial transcriptomics deconvolution and visualization to locate cells in the bone marrow tissue of three patients with AA. An anchor-based integration pipeline in Seurat was used to integrate the combined scRNA-seq dataset with 10x Visium spatial transcriptomics data. It enables the transfer of cell type annotations from single-cell RNA sequencing (scRNA-seq) to spatial transcriptomics. The cell type predictions in Seurat are imported into the R package SPOTlight, which provides annotations and visualizations of the cell types for each spatial location [26, 27].
Construction and grouping of AA mouse models. A total of ten C57BL/6 (B6) mice aged 213 days and 86 CB6FI mice aged eight weeks were purchased from Beijing Vitonglihua Experimental Animal Technology Co., Ltd., located in Beijing, China. The mice weigh between 20–25 g and are housed in standard cages. They are subject to a 12-hour light/dark cycle that alternates regularly, and the temperature in the room is kept constant at 23 ± 1°C. The mice have unrestricted access to food and water. Prior to commencing the experiment, the subjects will undergo a one-week acclimation period during which they will be exposed to adaptive feeding practices. The Institutional Animal Ethics Committee has approved this experiment and the animal use protocol(NO. 202309001).
First, we randomly divided the 86 CB6FI mice into two groups: the Control group, which comprised 16 mice, and the Model group, which included 70 mice. The construction method for the Model group AA mouse model is as described previously [28]. In brief, we retrieved the inguinal, axillary, and brachial lymph nodes from 10 C57BL/6 (B6) donor mice. The lymph nodes were then homogenized in Iscove's modified Dulbecco's medium (IMDM, ThermoFisher, 21056023) using a tissue grinder (Beyotime, E6600). Subsequently, they underwent washing, centrifugation, and filtration through a 70µm nylon mesh sieve (Labgic, 352350). The quantity of samples was determined using a Beckman Vi-Cell counter (Vi-CELL XR, USA). Subsequently, we will intravenously administer lymphocytes isolated from C57BL/6 (B6) donors, at a dosage of 5×106 lymphocytes per 200 µl of PBS, into CB6FI mice matched in gender. The CB6FI receptor mice underwent 5.0 Gy total body irradiation (TBI) 2–4 hours ago. A peripheral blood cell count was conducted to confirm the presence of symptoms associated with leukopenia, thereby confirming the successful construction of the model [29, 30]. Control group: intraperitoneal injection of PBS. Model group: The AA model was constructed, and intraperitoneal injection of PBS was performed.
The Control group will be randomly divided into two subgroups: The control group (8 mice, injected with PBS only) and the Control + NC-OE group (8 mice, injected with adenovirus empty vector NC-OE, which serves as the negative control for overexpression, along with PBS). Similarly, the Model group will be randomly divided into the following six subgroups: Model group (24 mice, undergoing modeling and injected with PBS), Model + Irisin group (14 mice, undergoing modeling and injected with Irisin), Model + NC-OE group (8 mice, undergoing modeling and injected with PBS and NC-OE), Model + Irisin + NC-OE group (8 mice, undergoing modeling and injected with Irisin and NC-OE), Model + Mst1/2-OE group (8 mice, undergoing modeling and injected with PBS and adenovirus vector Mst1/2 overexpression Mst1/2-OE), Model + Irisin + Mst1/2-OE group (8 mice, undergoing modeling and injected with Irisin and Mst1/2-OE).
The control group received no additional treatment except for an intraperitoneal injection of PBS. Model group: Only conducting AA modeling and PBS processing. In the Model + Irisin group, Irisin (MCE, HY-P72534) was dissolved in PBS and injected intraperitoneally at a dose of 100 µg/kg per injection, once a day for a total of 14 days, starting on the 4th day after constructing the AA model [31, 32, 33].
Furthermore, the virus transduction was conducted in the following manner: HEK-293 cells (Procell, CL-0001) were transfected with either the adenoviral vector (Mst1/2-OE, AAV-Mst1-Mst1) carrying the mouse Mst1 and Mst2 genes, or the adenoviral empty vector (NC-OE) plasmid. The transfection was performed using the LipoFiter transfection reagent (Hanbio, HB-LF-1000). Hanbio synthesized the plasmids. After 72 hours, collect the clear liquid from the top to obtain the viral fluid. After successfully constructing the AA model, different groups of mice were injected with Mst1/2-OE or NC-OE vectors via the tail vein 4 days later. The injection was performed with a concentration of 3.5×1012 viral genomes per mouse. Control + NC-OE group: Only tail vein injection of NC-OE and intraperitoneal injection of PBS were performed without constructing the AA model. Experimental Group: Four days after successfully constructing the AA model, NC-OE was administered through the tail vein while PBS was injected into the peritoneal cavity. Model + Irisin + NC-OE group: Accept model construction, NC-OE, and Irisin processing. Model + Mst1/2-OE group: After successfully establishing the AA model for 4 days, Mst1/2-OE was administered through a tail vein injection, followed by an intraperitoneal injection of PBS. Model: Mice treated with both Irisin and Mst1/2-OE were used to establish the AA model [34, 35].
Except for the 24 mice in the Model group and the 14 CB6FI mice in the Model + Irisin group, eight mice were randomly assigned to each of the remaining groups. On the second day following the administration of Irisin, all the mice were euthanized. Six mice from the Model group and six from the Model + Irisin group were used for transcriptome sequencing. Ten mice from the Model group were also used to extract bone marrow mesenchymal stem cells (BMSCs) for in vitro mechanistic validation.
Bone marrow smear. After decapitating the mouse, expeditiously remove the sternum and utilize hemostatic forceps to grip it, facilitating the extraction of the bone marrow. Gently place the bone marrow smear onto a microscope slide pre-treated with 0.05 mL of fetal bovine serum (FBS, Hyclone), then slide it forward. The identical procedure is employed for the bone marrow obtained from normal donors and bone marrow from patients with AA, which is collected through bonemarrow aspiration (BM aspiration). Next, allow the slides to dry at room temperature and position them on a staining rack. The cells were fixed by applying a drop of methanol for 3 minutes. Then, the working solution of Wright Giemsa (Solarbio, G1021) was added. The working solution was prepared by diluting 1 part of the stock solution with 9 parts of the buffer solution. The working solution allowed the cells to be completely covered at room temperature for 20 minutes. Thoroughly rinse the glass slides with distilled water from one end to the other. The examination and observation were conducted using an OLYMPUS BX46 upright microscope. The ImageJ software was used to quantify the percentage of nucleated cells in the bone marrow. Each group has three independent perspectives [31].
Pathological sections of femoral bone marrow. Mouse femurs were harvested, or normal donor and AA patient bone marrow were obtained through bone marrow biopsy (BM biopsy). The samples were fixed in 4% paraformaldehyde (Biosharp, BL539A) at room temperature for 48 hours. Then, rinse thrice with PBS and distilled water, each time for 20 minutes. Replace the EDTA solution (Solarbio, E1171) for decalcification every 7 days, totaling 28 days. After decalcification, rinse the femur or bone marrow using running water for 20 minutes in an embedding box. Subsequently, perform dehydration by exposing the sample to ethanol. After complete dehydration, clean with xylene. Embed the transparent mouse femur in paraffin. To obtain 4 µm thick sections, the paraffin blocks embedded in mouse femurs were sliced using a wax microtome (Leica, LECIA RM2235). Spread the sliced organisms onto a distillation water bath heated to 45°C. Then, transfer the slices onto clean glass slides, allowing them to drain and dry on a glass slide heater set at 65°C for one hour. De-waxing and dyeing procedures were carried out using an automated staining machine, with the dyeing time determined according to the manufacturer's instructions of the HE Staining Kit (Solarbio, G1120). Finally, seal the slices with neutral adhesive. Examine the bone marrow sections stained with hematoxylin and eosin under an inverted microscope (Leica, Leica DM IL LED) and photograph them at a magnification 200×. The nucleus appears blue after staining, while red blood cells and bone trabeculae are stained with eosin. Hematopoietic tissues comprise granulocytes, red blood cells, megakaryocytes, and lymphocytes, whereas non-hematopoietic tissues comprise trabecular bone, cortical bone, and adipose tissue [31]. ImageJ software was employed to identify and calculate the percentage area of hematopoietic tissue in the entire bone marrow interstitium. Three independent fields were quantified for each group.
Peripheral blood cell count. Mouse tail vein blood was collected on the 5th, 10th, and 15th days after establishing the AA mouse model to determine the absolute values of white blood cells (WBC), neutrophils (NEU), platelets (PLT), red blood cells (RBC), and hemoglobin (HGB) concentration in peripheral blood. It was done using a hematology analyzer (HORIBA, ABX PENTRA XL 80) [33].
The hematopoietic colony forms experiment. Following the death of the mice, a semi-solid culture system employing colony-forming units (CFUs), including CFU-erythroid (CFU-E), CFU-granulocyte macrophage (CFU-GM), and CFU-megakaryocytic (CFU-MK), was utilized to extract nucleated cells from the bone marrow of the femur and subsequently adhere them onto the wells of the tissue culture plates. Bone marrow cells were cultured in discover-modified Dulbecco's medium (IMDM, Sigma-Aldrich, I3390). The culture medium consisted of 20% fetal bovine serum (FBS, Gibco, USA), 300 mg/L glutamine (Sigma-Aldrich, G7513), and either 10 µg/L recombinant mouse macrophage colony-stimulating factor (GM-CSF, Sigma-Aldrich, M9170) or erythropoietin (EPO, ACROBiosystems, EPO-H4214). Additionally, a viscosity carrier of 0.3% agarose was included. Cells were cultured in Iscove's Modified Dulbecco's Medium (IMDM) for the CFU-MK culture. The medium was composed of 20% fetal bovine serum, 300 mg/L of glutamine, 1% bovine serum albumin (Sigma-Aldrich, A1933), 10 − 5 mol/L of 2-mercaptoethanol (Sigma-Aldrich, M3148), and 10 µg/L of recombinant mouse thrombopoietin (Sigma-Aldrich, T4184). The experiment was repeated three times, with 105 nucleated cells per well, and cultured in a humid environment at 37°C with 5% CO2. Following a 5-day cultivation period, the colonies of CFU-E (≥ 8 cells) and CFU-GM (≥ 40 cells) were enumerated. As previously stated, mouse megakaryocyte cells in CFU-MK colonies were detected using acetylcholinesterase staining. The count of CFU-MK colonies with four or more cells was performed after seven days of culture [33].
Organize slice oil red o staining. Following decalcification and dehydration, the femur sections of the mouse were stained with a mixture of 0.21% Oil Red O (Sigma, 1320-06-5) and 100% isopropanol (Sigma, 67-63-0) for 10 minutes. The images were captured using the Olympus BX53 microscope, and the percentage of positive area of fat cells was quantified using ImageJ software. Three mice are randomly selected for each group, and three sections are extracted from each mouse. Data from three independent fields of view are analyzed for every section [28].
Immunofluorescence staining. Following dewaxing and dehydration, paraffin sections of the mouse femur underwent antigen retrieval at 98°C. Incubation was then carried out using 1% Triton X-100 (Sigma-Aldrich, X100) in PBS, followed by a 30-minute blocking step with goat serum (Solarbio). The BMSCs cells were fixed with 4% paraformaldehyde (Biosharp, BL539A) for 10 minutes, permeabilized with 0.1% Triton X-100 for 15 minutes, and then blocked with 1% BSA (Thermofisher, 37520) at room temperature for 1 hour. Mouse femoral bone slices and BMSCs cells were subjected to immunofluorescent staining using Perilipin antibody (1:200, Thermofisher, MA5-32597) or YAP antibody (1:500, Thermofisher, PA1-46189). The antibodies were incubated overnight at 4℃, followed by a 40-minute incubation at room temperature with Alexa FluorTM Plus 488 goat anti-rabbit IgG antibody (1:1000, Thermofisher, A32731). The samples were stained with 4',6-Diamidino-2-Phenylindole (DAPI) at 0.5µg/mL (Invitrogen, D3571). Subsequently, the stained samples were observed, and images were captured using a fluorescence microscope (Olympus FV-1000/ES). The data were analyzed using the ImageJ software. Three mice were randomly selected for each group. Three slices were taken from each mouse. The fluorescence intensity was assessed in three independent fields of view for each slice [36, 37].
Cell culture. Bone marrow mesenchymal stem cells (BMSCs) were isolated from the bone marrow of mice by flushing the femurs and tibias of both wild-type mice and mice from different treatment groups after decapitation. The centrifugation was performed at 1000 rpm for 5 minutes. The suspension of bone marrow cells was cultured in L-DMEM (Thermo Fisher, A1443001, USA), supplemented with 10% fetal bovine serum (Gibco, USA) and 1% penicillin/streptomycin. The culture was carried out in a 10 cm² culture dish. The culture medium is changed once a day for three consecutive days to remove non-adherent cells. Afterward, change the culture medium every 3 days.
Additionally, to isolate human BMSCs, three-milliliter bone marrow samples were collected from patients with AA and normal donors. Bone marrow mononuclear cells were separated from these samples using Ficoll-Paque (Cytiva, 17144003, USA) through centrifugation. Subsequently, these cells were seeded in 10 cm2 culture dishes, utilizing the previously mentioned culture medium. Remove unattached cells after 48 hours and replace the culture medium every 3 days afterward. Cultured adherent BMSCs with a cell density of 80%-90% were subjected to trypsin treatment. Isolated BMSCs were characterized using flow cytometry [28, 38].
Flow cytometry. Surface marker analysis was conducted on BMSCs obtained from normal donors, patients with AA, and mouse bone marrow. FITC-conjugated anti-human CD45, HLA-DR, CD34, CD105, CD44, and CD29 antibodies were used for human samples, while anti-mouse CD45, HLA-DR, CD34, CD105, CD44, and CD29 antibodies were used for mouse samples. After incubating the cells at 4°C in the dark for 30 minutes, wash the BMSCs twice with PBS and then centrifuge them at 2000 rpm for 5 minutes at 4°C. We analyzed the percentage of cells expressing these surface markers using the FACSCanto II flow cytometer (BD, USA) [39, 40].
Induction of adipogenic or osteogenic differentiation in BMSCs and cellular grouping. Bone marrow mesenchymal stem cells (BMSCs) were seeded at a density of 2.5×106 cells per well in a six-well plate. For differentiation induction, the cells were cultured in a medium comprising 10% fetal bovine serum (FBS), 1% penicillin/streptomycin, 0.5 mM 3-isobutyl-1-methylxanthine (Sigma-Aldrich, I7018), 1 µM dexamethasone (Sigma-Aldrich, D4902), and 5 µg/mL insulin (Sigma-Aldrich, I3536). The medium for inducing adipogenic differentiation is changed every 3 days, and the induction process is carried out for 14 days [38].
Bone marrow-derived mesenchymal stem cells (BMSCs) were seeded at a density of 2 × 106 cells per well in a 6-well plate. Once the cells reached 80% confluence, they were cultured in osteogenic induction medium comprising 10% fetal bovine serum (FBS), 1% penicillin/streptomycin, 0.1 mM dexamethasone (Sigma-Aldrich, D4902), 10 mM β-glycerophosphate (Sigma-Aldrich, G9422), and 50 mM ascorbic acid (Sigma-Aldrich, A4403). The osteogenic induction medium is changed every 3 days [32, 38].
The cell groups involved in the study were divided into five categories. These included the Normal group, AA group, Control group, Model group, and Model + Irisin group. The groups consisted of BMSCs isolated from clinical samples or different treatment groups of mice. BMSCs were differentiated from clinical normal donor bone marrow samples in the Normal group. BMSCs were differentiated from patient bone marrow samples with clinical AA in the AA group. In the Control group, BMSCs were differentiated from mouse bone marrow samples. In the Model group, BMSCs were differentiated from mouse bone marrow samples. In the Model + Irisin group, BMSCs were differentiated from mouse bone marrow samples with Model + Irisin treatment.
Next, BMSCs cells from the Model group mice will be divided into 24 treatment groups, including the PBS group, DMSO group, PBS + DMSO group, Irisin group, Irisin + DMSO group, VP group, PBS + VP group, Irisin + VP group, Rotenone group, VP + Rotenone group, NC-OE group, Mst1/2-OE group, Yap-OE group, Mst1/2-OE + Yap-OE group, PBS + NC-OE group, PBS + DMSO + NC-OE group, Irisin + NC-OE group, Irisin + DMSO + NC-OE group, Irisin + VP + NC-OE group, PBS + Mst1/2-OE group, Irisin + Mst1/2-OE group, Irisin + DMSO + Mst1/2-OE group, PBS + Yap-OE group, and Irisin + Yap-OE group. The differentiation-inducing medium was supplemented with PBS, DMSO, Irisin, VP (verteporfin), and Rotenone. These drugs were co-treated with BMSCs, and the duration of drug treatment coincided with the induction time.
In the PBS group, DMSO group, and PBS + DMSO group, the differentiation-inducing culture medium was supplemented with PBS, DMSO, or PBS + DMSO, respectively, to induce the differentiation of BMSCs. The irisin group involved dissolving irisin in PBS at 1, 5, 10, and 20 ng/ml concentrations. The dissolved irisin was then added to the differentiation-inducing culture medium to co-treat BMSCs and determine the optimal concentration. In the Irisin + DMSO group, Irisin and PBS were co-administered to treat BMSCs and induce differentiation. In the VP Group, add verteporfin (verteporfin, VP; MCE, HY-B0146) dissolved in DMSO at a concentration of 0.8 µM to the differentiation-inducing medium for simultaneous treatment with BMSCs. PBS and VP collaborate to process BMSCs and prompt differentiation. In the Irisin + VP group, both Irisin and VP were included in the differentiation-inducing medium to co-treat BMSCs. In the Rotenone group, a concentration of 100 nM of the fish poison Rotenone (Rotenone, MCE, HY-B1756) dissolved in DMSO was added to the differentiation induction medium to be co-treated with BMSCs. VP + Rotenone group incorporated VP and Rotenone into the differentiation-inducing culture medium to co-treat with BMSCs [41, 42, 43].
Furthermore, the AAV-Yap1 plasmid containing the mouse Yap1 gene, synthesized by Hanheng Biotechnology, will be transfected into HEK-293 cells after packaging. It will be done following the animal grouping section, along with the Mst1/2-OE vector and NC-OE vector viral solutions for transfecting BMSCs cells. BMSCs cells in the logarithmic phase (5 × 104 cells/well) were seeded in a 24-well plate. Transfection occurred when the cell confluence reached 50%, with a multiplicity of infection of 5 × 104 vp/cell. After 24 hours, the culture medium was replaced with an L-DMEM medium containing 10% FBS, and the medium was changed every three days [44].
The cell treatment after 3 days of viral transduction is as follows: In the NC-OE group, differentiation induction occurs after receiving NC-OE transfection. In the Mst1/2-OE group, differentiation induction was performed following transfection with Mst1/2 overexpression. Yap-OE group: Differentiation induction after Yap-OE transfection. In the Mst1/2-OE + Yap-OE group, differentiation induction was conducted following the co-transfection of Mst1/2-OE and Yap-OE. The PBS + NC-OE group was treated with PBS and NC-OE to induce differentiation. The PBS + DMSO + NC-OE group was treated with PBS, DMSO, and NC-OE to induce differentiation. The Irisin + NC-OE group received treatment with Irisin and NC-OE to induce differentiation. In the Irisin + DMSO + NC-OE group, the participants were treated with Irisin, DMSO, and NC-OE, which led to induced differentiation. The Irisin + VP + NC-OE group consisted of subjects who received Irisin, VP, and NC-OE treatments to induce differentiation. In the PBS + Mst1/2-OE group, cells were initially treated with PBS and then induced to differentiate by overexpressing Mst1/2 proteins. The Irisin + Mst1/2-OE group received treatment with Irisin and Mst1/2-OE to induce differentiation. Irisin + DMSO + Mst1/2-OE group: The individuals in this group received treatment with Irisin, DMSO, and Mst1/2-OE and were induced for differentiation. The PBS + Yap-OE group accepted both PBS and Yap-OE treatments to induce differentiation. The Irisin + Yap-OE group was treated with Irisin and Yap-OE to induce differentiation [44, 45].
Oil red O staining of BMSCs. After 14 days of induction for adipogenesis, the cells were fixed in a 4% paraformaldehyde solution (Biosharp, BL539A) and stained with Oil Red O solution (Sigma-Aldrich, MAK194). The dye should be dissolved in isopropanol, and the absorbance at 510 nm should be measured [38].
Staining of alkaline phosphatase (ALP) and alizarin red was performed in BMSCs. On the seventh day of osteogenic induction, BMSCs were fixed using a 4% paraformaldehyde solution (Biosharp, BL539A) for 30 minutes. Subsequently, they were covered with a BCIP/NBT working solution (Beyotime, C3206) and incubated in the dark for 20 minutes. Observe cells under a microscope and take pictures. The protein concentration was measured using the BCA Protein Quantification Kit (ThermoFisher, 23227), and the OD values of the blank wells, standard wells, and test wells were determined at 520 nm using the alkaline phosphatase (ALP) staining kit (Jiancheng Bioengineering Research Institute, A059-2-2). Subsequently, the alkaline phosphatase activity was calculated following instructions [32].
After 21 days of osteogenic induction, the cells were fixed using 4% paraformaldehyde (Biosharp, BL539A) and stained with 2% Alizarin Red S (Sigma-Aldrich, A5533). The Xanthein Red S should be dissolved in a solution of hexadecylpyridinium chloride (Sigma-Aldrich, 1104006) and then quantitated using spectrophotometry at a wavelength of 562nm [32].
Western blot. Bone marrow stromal cells (BMSCs) were lysed using the RIPA total protein lysis buffer (AS1004, Wuhan Aspen Biotechnology Co., Ltd., China) after 7 days of osteogenic or adipogenic differentiation. The protein concentration was subsequently measured using the BCA protein quantification assay kit (23227, Thermo Fisher).
Proteins were separated using SDS-PAGE and then transferred onto a PVDF membrane. The membrane was blocked with 5% BSA at room temperature for 1 hour. Afterwards, the primary antibodies were added individually, including LPL (...) Incubate the antibody at 4℃ overnight.
LPL, FABP4, PPARγ, CEBPα, and PERILIPIN are markers for adipogenic differentiation, whereas Runx2, ALP, OPN, and OCN are markers for osteogenic differentiation. MST1, MST2, YAP, and p-YAP are proteins associated with the MST1/2-YAP signaling pathway.
The membrane was washed three times with TBST (3×5 minutes), followed by incubation with Anti-Mouse-HRP secondary antibody (1:10000; ThermoFisher, 31430) or Goat anti-Rabbit-HRP secondary antibody (1:10000; ThermoFisher, 31460) at room temperature for 2 hours. The membrane was washed thrice with TBST (3×5 minutes). TBST should be replaced with an appropriate ECL working solution (Millipore, WBKLS0500). The transfer membrane should be incubated at room temperature for 1 minute. Excess ECL reagent should be removed, and the membrane should be sealed with cling film. Before development and fixation, the membrane should be placed in a dark box for 5–10 minutes to expose an X-ray film. The ImageJ analysis software was used to quantify the grayscale intensity of bands in Western blot images, utilizing GAPDH as the internal reference [39].
RT-qPCR. The Trizol Reagent kit (Invitrogen, 10296028CN) should be utilized for cell lysis and total RNA extraction from cells or bone marrow tissue. UV-Vis spectrophotometry (ND-1000, Nanodrop, USA) was employed to evaluate RNA quality and concentration.
The PrimeScript™ RT-qPCR Kit (TaKaRa, RR037Q) was employed to measure mRNA expression levels for reverse transcription. Real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) was performed using SYBR Premix Ex TaqTM (TaKaRa, RR390A) on a LightCycler 480 system (Roche Diagnostics, Pleasanton, CA, USA). GAPDH is a reference gene that functions as an internal control for mRNA. The primers utilized for amplification were designed and provided by Shanghai Universal Biotech Co., Ltd. The primer sequences can be found in Table S1. The term 2−ΔΔCt represents the fold difference in the target gene expression between the experimental and control groups. The formula is defined as ΔΔCT = ΔCt experimental group - ΔCt control group, where ΔCt is calculated as the difference between the Ct values of the target gene and reference gene [46].
RNA extraction, library construction, and sequencing. Total RNA was extracted from bone marrow stromal cells (BMSCs) and isolated from two groups of mice: the Model group (n = 6) and the Model + Irisin group (n = 6). The extraction was performed using a Trizol reagent (15596026, Invitrogen, USA). The concentration and purity of RNA samples were determined using a spectrophotometer instrument, specifically the Nanodrop 2000 (1011U, Nanodrop, USA). Total RNA samples meeting the following criteria are utilized for subsequent experiments: RNA Integrity Number (RIN) ≥ 7.0 and 28S:18S ratio ≥ 1.5 [47].
CapitalBio Technology, located in Beijing, China, generated and sequenced the sequencing library. Every sample utilizes a total of 5 µg of RNA. We employed the Ribo-Zero Magnetic Kit (MRZE706, Epicentre Technologies) to remove ribosomal RNA (rRNA) from total RNA. The NEB Next Ultra RNA Library Prep Kit (#E7775, New England Biolabs, USA) generates Illumina-compatible sequencing libraries. Next, the RNA fragments were fragmented into 300 base pairs (bp) using the NEB Next First Strand Synthesis Reaction Buffer (5×). The first-strand cDNA is synthesized using a reverse transcriptase primer and random primer, while the second-strand cDNA is synthesized in the reaction buffer of dUTP Mix (10×) for the second-strand synthesis. The repair of cDNA fragments involves adding polyA tails and connecting sequencing adaptors. Following the ligation of Illumina sequencing adapters, the second cDNA strand was digested using the USER enzyme (#M5508, NEB, USA) to generate strand-specific libraries. The library DNA should be amplified, purified, and enriched through PCR. Next, the libraries were evaluated using the Agilent 2100 system and quantified using the KAPA Library Quantification Kit (KK4844, KAPA Biosystems). Lastly, we conducted paired-end sequencing using the Illumina NextSeq CN500 sequencer [48, 49].
Quality control of sequencing data and its alignment to a reference genome. The quality of the paired-end reads in the raw sequencing data was assessed using FastQC software version 0.11.8. The raw data underwent preprocessing using Cutadapt software version 1.18, which involved removing Illumina sequencing adapters and poly(A) tail sequences. Filter out reads with an N content exceeding 5% using a perl script. Using the FASTX Toolkit software version 0.0.13, we extracted reads with a base quality of 20 or higher, which accounted for 70% of the total. Repair the paired-end sequences using BBMap software. Finally, the filtered fragments of high-quality reads were aligned to the mouse reference genome using hisat2 software (version 0.7.12) [50, 51].
Differential expression gene bioinformatics analysis. The limma package in R was used to identify differentially expressed genes (DEGs) in the raw count matrix. DEGs were selected based on a threshold of |log fold change (FC)| > 1 and a P value < 0.05. The ggplot2 package in R was used to plot the volcano plot. The clustering heat map of differentially expressed genes (DEGs) was generated using the "pheatmap" package in the R programming language. We conducted enrichment analysis using the R language and several packages, including the "clusterProfiler", "org.Hs.eg.db", "org.Mm.eg.db", "enrichplot", "ggplot2", and "pathview" packages. The analysis focused on Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Gene Set Enrichment Analysis (GSEA) [52].
Bioinformatics analysis. We extracted 26 mouse genes related to the Hippo signaling pathway (HRGs) from the Reactome database (https://reactome.org/). Conduct a protein-protein interaction (PPI) analysis of HRGs using the STRING database (https://string-db.org/) and optimize the results using Cytoscape 3.10.0 software. The Jvenn website generates Venn diagrams to identify the intersection of differentially expressed genes (HR-DEGs) associated with the Hippo pathway. The clustering heatmap, volcano plot, correlation analysis heatmap, and scatter plot for highly-regulated differentially expressed genes (HR-DEGs) were generated using the R packages "ggplot2" and "pheatmap" [53].
GO and KEGG are utilized to perform functional enrichment analysis of HR-DEGs. GO and KEGG analyses were conducted using the 'clusterProfiler,' 'org.Hs.eg.db,' 'enrichplot,' 'DOSE,' and 'ggplot2' packages in the R programming language [54].
We applied machine learning algorithms such as lasso regression, SVM-RFE, and random forest using the "glmnet", "e1071", and "randomForest" packages in the R programming language. The Venn diagram was plotted using the "venn" package in R [55].
CCK-8. Cell viability was assessed following the guidelines provided in the CCK-8 assay kit (ab228554, Abcam, USA). Cells from each group were seeded into individual wells of 96-well plates (2500 cells per well) and cultured for 1, 3, and 5 days, respectively. At the designated time point, add 10µl of CCK-8 to each well. After incubating for 2 hours, the absorbance at 450 nm should be measured using an enzyme-linked immunosorbent assay reader (M1000 PRO, Tecan) [41].
Live-death staining. The LIVE/DEAD Cell Viability/Cytotoxicity Assay Kit (Invitrogen, L3224) was employed to assess cell death objectively. The assay kit offers two types of molecular probes: one that labels live cells as green based on intracellular esterase activity and another that labels dead cells red due to compromised membrane integrity. Conduct the testing in accordance with the plan provided by the manufacturer. In brief, cells were seeded in a 24-well plate and incubated overnight. Afterward, they were treated with Irisin for 3 days. Subsequently, the cells were incubated with a fluorescent dye (2.0 µM) for 15 minutes, and a fluorescence microscope (FV-1000/ES, Olympus) was used to capture microscopic images. The software ImageJ was employed to identify and calculate the percentage of viable and nonviable cells [56].
MitoTracker staining. After inducing adipogenic differentiation of BMSCs for 14 days, the mitochondria staining reagent, mitoTracker Green FM (Invitrogen, M7514), was applied at a concentration of 100 nM. The cells were then incubated in a cell culture incubator for 30 minutes. Next, the Hoechst 33342 Live Cell Stain (Beyotime, C1029) should be applied in a dark environment at 37°C for 10 minutes. Subsequently, the cells were washed twice with preheated PBS, and a fresh culture medium was added. Imaging was then promptly conducted using a Zeiss LSM 510 META confocal microscope (Zeiss). Merge and scale the original image using ImageJ software [43, 57].
Quantitative analysis of mitochondrial DNA. Total genomic and mitochondrial DNA could be extracted using the QIAamp DNA Mini kit (Qiagen, 51304), following the manufacturer's instructions. Adjust the DNA template concentration to 10 ng/µl. Refer to the report by Malik et al. [58]. Evaluating the amount of mitochondrial DNA through RT-qPCR for absolute quantification. We utilized mouse mitochondrial DNA primers (Mito) and mouse nuclear DNA primers (β2-microglobulin, mB2M) to amplify the corresponding products in mouse genomic DNA. The primer sequences are available in Table S2 [59].
Measurement of the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR). The Seahorse XF24 extracellular flux analyzer (Agilent, Seahorse XF24, 24-well plate) was employed, following the described method, to assess the oxygen consumption rate (OCR) of adipogenic-induced BMSCs after 7 days [60, 61]. The OCR determination was conducted using the Seahorse XF Cell Mito Stress Test Kit (Agilent, 103672-100). The assay medium consisted of Seahorse XF DMEM culture medium (Agilent, 103680-100) supplemented with 1 mM pyruvate, 2 mM glutamine, and 10 mM glucose, adjusted to pH 7.4. Sixty thousand cells were inoculated per well (0.32 cm2 growth area) into XF24 24-well culture microplates containing 500 µL of assay medium. The plates were incubated overnight in a humid environment at 37℃, with 95% air and 5% carbon dioxide. The culture medium should be removed before the experiment, and 500 µL of fresh assay medium should be added. The cells are pre-incubated in ambient air at 37°C for one hour. We used oligomycin at a concentration of 4 µg/mL to assess ATP synthesis driven by oxidative phosphorylation and respiration driven by proton leak. Following three measurement cycles, a decoupling agent, carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP), was added at a concentration of 5 µM to assess the maximum respiratory capacity. After three additional measurement cycles, 1 µM of Rotenone should be added to inhibit complex I and 1 µM of antimycin A should be used to inhibit complex III, thereby suppressing mitochondrial respiration.
Furthermore, the Seahorse XF Glycolysis Stress Test Kit (Agilent, 103020-100) was used to accurately measure the glycolytic rate. The cells should be replaced with XF DMEM medium (Agilent, 103575-100) supplemented with 5 mM HEPES at pH 7.4. Additionally, 1 mM pyruvate (Agilent, 103578-100), 2 mM glutamine (Agilent, 103579-100), and 10 mM glucose (Agilent, 103577-100) should be added. The cells should then be incubated in a CO2-free incubator at 37℃ for 1 hour. Before commencing the test, the culture medium should be replaced again, following the guidelines provided by the manufacturer. After establishing the baseline, 1 µm fisetin and antifungal A were added sequentially, along with 50 mM 2-deoxyglucose, and the response was measured. The OCR, ECAR, glycolytic proton efflux rate, and ATP production rate were determined using the Seahorse XFe96 software, version 2.6 [62].
ROS detection. The levels of reactive oxygen species (ROS) inside the cells were measured in mouse bone marrow-derived mesenchymal stem cells (BMSCs) or BMSCs induced for adipogenic differentiation for 7 days from different groups, following the guidelines provided by the manufacturer. The measurement was performed using a ROS assay kit (Beyotime, S0033S). The fluorescent probe DCFH-DA (10 mM) should be diluted 1000-fold in serum-free L-DMEM before adding it to the cells. The cells should be incubated in the dark at 37℃ for 20 minutes. After rinsing with PBS three times, we examined cell morphology using a fluorescence microscope (FV-1000/ES, Olympus). Subsequently, the cells were harvested, and the fluorescence intensity was quantified using a TriStar3 multimode reader (Berthold Technologies) with an excitation wavelength set at 488 nm and an emission wavelength of 525 nm. Normalize intracellular ROS levels to total cell number [63].
Mitochondrial membrane potential (MMP) detection. After isolating bone marrow-derived mesenchymal stem cells (BMSCs) or differentiating them into adipocytes for 7 days, we assessed the mitochondrial membrane potential (MMP) using the JC-1 Mitochondrial Membrane Potential Assay Kit (Beyotime, C2006) [64]. In summary, the cells are incubated in a culture medium containing JC-1 for 30 minutes. Subsequently, they are loaded into the BD FACSMelody flow cytometer (BD) with an excitation wavelength of 480 nm and emission wavelengths of 525 nm and 590 nm [65]. We analyzed 10,000 cells using FlowJo X10 software and repeated the experiment three times.
ATP level determination. The total ATP production of BMSCs in each group was quantified using the ATP assay kit (Beyotime, S0026B). Initially, the cells are inoculated into a 96-well plate. Subsequently, PBS washing, lysis, and centrifugation are performed. Subsequently, the luminescence in the supernatant was measured with the BioTek Synergy 2 microplate reader (BioTek Instruments Inc.). The MicroBCA Protein Assay Kit (ThermoFisher, 23235) was utilized to determine the protein concentration and subsequently performed normalization. Finally, the ATP content should be determined using the ATP standard curve, and the experiment should be repeated three times [66].
Statistical analysis. Each experiment was repeated independently at least three times, and the data are presented as the mean ± standard deviation (SD). To compare the differences between groups, we utilize either an independent samples t-test or a one-way analysis of variance. If the variance analysis results reveal differences, we will conduct Tukey's HSD post-hoc test to examine the disparities between each group. When dealing with data that is not normally distributed or exhibits heteroscedasticity, we will employ either the Mann-Whitney U or Kruskal-Wallis H test. Statistical analyses were conducted using GraphPad Prism 8.0 software [28]. A p-value less than 0.05 is considered statistically significant.