Haploinsufficiency at the CX3CR1 locus of hematopoietic stem cells favors the appearance of microglia-like cells in the central nervous system of transplant recipients

Mouse studies

All experiments and procedures involving animals were performed with the approval of the Dana Farber Animal Facility Institutional Animal Care and Use Committee (IACUC 15-031). Mice have been housed in appropriate rooms with controlled temperature, humidity, and automated lighting, ensuring a standard light/dark circadian cycle. Wild-type C57BL/6J mice or C57BL6/Ly5.1 mice (hereafter called CD45.2 or CD45.1, respectively) were obtained from Jackson Lab or Charles River. NOD-SCID-IL2Rg−/− (NSG) female mice (Stock No: 005557) were obtained from The Jackson Laboratory.

Cx3cr1^GFP/+ (referred to as Cx3cr1^−/+) mice were generated by crossing Cx3cr1^GFP/GFP (referred to as Cx3cr1^−/−) obtained from The Jackson Laboratory (Stock. No 005582) with wild type CD45.2 mice (referred as Cx3cr1^+/+). For all transplantation experiments, mice were randomly distributed to each experimental group. Female animals were used as recipients due to their better response to busulfan conditioning and homogeneity in the engraftment. Female and male mice were used as donors.

Isolation, transduction, and transplantation of murine hematopoietic cells

Seven/eight-week-old wild type, Cx3cr1^−/+ or Cx3cr1^−/− mice were euthanized with CO₂, and the BM was harvested by crushing bones. After BM lysis, HSPCs were purified by Lin- selection using the Lineage Cell Depletion Kit (Miltenyi, #130-090-858) with the autoMACS™ magnetic separation, following the manufacturer’s instruction.

Sorting experiments

GFP⁺ and GFP^– cells were sorted from the Lin- pool isolated from Cx3cr1^−/+ mice using the BD FACSAria II high-speed cell sorter. Collected cells were freshly transplanted IV into busulfan (4 doses 25 mg/kg) conditioned CD45.1 recipients at a 1:1 donor/recipient ratio. Mice also received 1.0 × 10⁶ CD45.1 BMNC IV 5 days post-transplant for hematopoietic rescue. Forty-five days post-transplant mice were sacrificed, and BM and brain were collected for cytofluorimetric analysis.

Standard transplantation experiments

Isolated Cx3cr1^+/+ or Cx3cr1^−/+ Lin- were transplanted IV into busulfan-conditioned CD45.1 recipients (1.0*10⁶/mouse) after 12–16 h of culture in StemSpan medium supplemented with cytokines as previously described⁵⁰. Mice were sacrificed at 45, 90, and 180 days post-transplant to collect hematopoietic organs and brain for flowcytometric analysis.

Competitive experiments

Isolated Cx3cr1^+/+, Cx3cr1^−/+, and Cx3cr1^−/− Lin- were transduced with LVs, for 12–16 h at multiplicity of infection (MOI) 100 ⁵¹. The following LVs were used:

pCCLsin.cPPT.humanPGK.BlueFluorescentProtein.Wpre (BFP-LV) for Cx3cr1^−/+ or Cx3cr1^−/−HSPCs and pCCLsin.cPPT.humanPGK.mCherryProtein.Wpre (mCherry-LV) for Cx3cr1^+/+ HSPCs. A fraction of the transduced cells was cultured for 10 days in vitro⁵⁰ to assess transgene expression by cytofluorimetric analysis. Transduced cells were injected via the tail vein or directly in the CNS by means of ICV injection into seven/eight-week-old conditioned CD45.1 female mice as previously described²⁰.

For IV group a total of 1.0 × 10⁶ cells/mouse (0.5 × 10⁶ Cx3cr1^−/+ or Cx3cr1^−/− BFP⁺ HSPCs + 0.5 × 10⁶ Cx3cr1^+/+ mCherry⁺ HSPCs) was injected. For ICV injection, a total of 0.3 × 10⁶cells/mouse (0.15 × 10⁶ Cx3cr1^−/+ BFP⁺ HSPCs + 0.15 × 10⁶ Cx3cr1^+/+ mCherry⁺ HSPCs) was injected. ICV transplanted mice received also 1.0 × 10⁶ CD45.1 BMNC IV 5-day post-transplant for hematopoietic rescue.

Trans-well migration assay

A trans-well migration assay was performed using HSPCs isolated from naïve wild type and Cx3cr1^−/+ mice. 2 × 10⁵ purified HSPCs were seeded in the upper chamber of 5-μm-pore filters trans-well plate, while in the bottom chamber Fractalkine (FLK), Cx3cr1 ligand, was added at different concentrations (100 and 200 ng/mL). The percentage of migrating cells was analyzed after 24 h. As a control, in another chamber SDF-1 (50 nM) was added, as HSPCs are expected to chemotactically migrate in the presence of this chemokine²⁵. Additionally, the same assay was performed with RAW 264.7, a murine macrophage cell line expressing high levels of Cx3cr1 and, therefore, expected to migrate in the presence of FLK.

Cell lines and primary cell culture

HEK293T cells used for AAV and LV productions were cultured in Dulbecco’s Modification of Eagle’s Medium (DMEM) 4.5 g/L glucose or Iscove’s modified Dulbecco’s medium (IMDM; Corning), respectively, supplemented with 10% heat-inactivated fetal bovine serum (FBS), 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2% glutamine. To choose and test the tools for gene editing at the CX3CR1 locus, we used RPMI 8226 (ATCC #CCL-155™), a suitable transfection host cell line expressing CX3CR1 gene, and K562 (ATCC #CCL-243™), not expressing CX3CR1, as negative control. RPMI 8226 were cultured in RPMI medium supplemented with 10% FBS, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2% glutamine. K562 were cultured in Iscove’s modified Dulbecco’s medium (IMDM; Corning) supplemented with 10% FBS, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2% glutamine. Both cell lines were cultured in a 5% CO₂ humidified atmosphere at 37 °C.

Commercial cord blood CD34+ human HSPCs were purchased frozen from AllCells and were thawed using thymocyte thawing media (TTM) medium, modified from ref. ⁵¹. TMM medium was prepared with RMPI medium supplemented with 30% FBS, 1%Pen/Strep, 10 ug/ml DNase I (Sigma), and 20 U/ml heparin. Cord blood CD34+ cells were gently thawed in pre-warmed TMM medium and left in the water bath at 37 °C for 1 h. Cells were then spun down (300×g 5 min) and seeded at the concentration of 5 × 10⁵ cells/ml in a cytokine-enriched medium for prestimulation. Serum-free StemSpan SFEM II medium (StemCell Technologies, #09605) was supplemented with 100 IU/ml penicillin, 100 μg/ml streptomycin, 100 ng/ml hSCF (PeproTech), 100 ng/ml hFlt3-L (PeproTech), hTPO 100 ng/mL (PeproTech), hIL-6 (PeproTech) and SR1 (StemRegenin1; 0.75 μM, CellagenTech). Cells were left in prestimulation for 2 days before the editing procedure. Cord blood CD34+ cells were cultured in a 5% CO₂ humidified atmosphere at 37 °C in low oxygen conditions.

LV and AAV6 production

Third-generation SIN Lentiviral vectors were produced and titered as previously described in ref. ⁵⁰. For in vitro experiments, AAV6 vectors were produced using the AAVpro® Purification Kit (Takara Bio #6666) according to manufacturer instructions and titered as previously described in ref. ³⁰. For in vivo experiments, AAV6 vectors were produced using iodixanol gradients and ultracentrifugation by the Boston Children’s viral vector core.

CRISPR/Cas9 Gene editing in cell lines and in human hematopoietic stem cells

Sequences of the gRNAs were designed using an online tool⁵² and selected for predicted specificity score and on-target activity. AAV6 donor templates were generated from a construct containing AAV2 inverted terminal repeats (ITRs) as previously reported²⁹. Vector maps were designed with SnapGene software v5.0.7 (from GSL Biotech; available at snapgene.com).

For experiments in cell lines, RNP complexes were assembled by incubating at 1:1.5 molar ratio Streptococcus pyogenes (Sp)Cas9 protein (integrated DNA technologies-IDT) with pre-annealed synthetic Alt-R® crRNA:tracrRNA (IDT) for 15’ at room temperature together with 0.1 nmol of Alt-R® Cas9 Electroporation Enhancer (IDT), added prior to electroporation according to manufacturer’s instructions. Both RPMI 8226 and K562 were nucleofected (FF120 program, Lonza 4D-Nucleofector) with the SF Cell 4 d Nucleo Kit (Lonza #V4XC-2032) using 50 pM of the different sgRNAs combined with CRISPR/Cas9 in the form of RNP complexes. AAV6 transduction (20,000MOI) was performed right after electroporation, maintaining the cells at a density of 0.25 × 10⁶ cells/ml. After 5 h, cells were diluted in the proper medium and moved to a bigger well to maintain a cell density of 0.5–0.7 × 10⁶ cells/ml overnight. The following day (~12–16 h after), cells were washed with PBS and seeded for expansion for molecular characterization.

For experiments in primary hematopoietic stem cells (CD34), RNP complexes were assembled by incubating at RT for 5′ (Sp)Cas9 protein (IDT) with synthetic single guide RNAs (sgRNA) chemically modified (with 2′-O-methyl at 3 first and last bases, 3′ phosphorothioate bonds between first 3 and last 2 bases and addition of 80-mer SpCas9 scaffold to create a single guide RNA) obtained from Synthego. Per 300,000–500,000 cells in in vitro experiments, nucleofection strips from LONZA Kit S (P3 Primary Cell 4D-Nucleofector™ X #V4XP-3032) were used, mixing 6 μg of Cas9 protein with 3.2 μg of sgRNA at room temperature for 5’. For in vivo experiments, the Lonza 4D nucleocuvette cuvettes from LONZA Kit L (P3 Primary Cell 4D-Nucleofector™ X #V4XP-3012) were used, scaling up the amounts of cells and reagents of five folds.

After thawing, cells were put in prestimulation for 2 days in the cytokine-enriched medium. Cells were then washed with PBS, counted, and nucleofected with the assembled RNP complexes (DZ100 program, Lonza 4D-Nucleofector). AAV6 transduction was performed right after electroporation at 20 K MOI, maintaining the cells at a density of 0.25 × 10⁶ cells/ml. After 5 h, cells were diluted with the cytokine-enriched medium and moved to a bigger well to maintain a cell density of 0.5–0.7 × 10⁶ cells/ml overnight. The following day (~12–16 h after), cells were washed with PBS and seeded for expansion for flowcytometric analysis and molecular characterization or were prepared for transplantation.

Molecular characterization of edited cells

TIDE analysis

Four to five days after electroporation, a fraction of cells (~10⁵) was collected. DNA was extracted using DNA Extraction QE Buffer (Biosearch Technologies #QE09050) and amplified using different primer assays according to the target genomic region. A list of primers and thermal protocols is available in supplementary material. After amplification, the PCR product was run on 1.5% Agarose gel (Tte Laboratories Inc., #BMAG01), and DNA was extracted from the gel with GeneJET Gel Extraction Kit (Life Technologies, # K0692). Purified DNA was then sequenced (SANGER seq) to evaluate the percentage of Insertions and Deletions (INDELs) with the TIDE (Tracking of Indels by Indels by DEcomposition) software⁵³ (Brinkman et al., 2014).

Targeted integration (HDR)

To evaluate the efficiency of integration by homology direct repair (HDR), DNA was extracted with the QIAamp DNA Micro Kit (QIAGEN) from cells expanded in vitro for 10–14 days. About 20–50 ng of genomic DNA were analyzed using the QX200 Droplet Digital PCR System (Bio-Rad) according to the manufacturer’s instructions. Different assays of primers and probes were designed on the junction between the vector sequence and the targeted locus. Human albumin was used for normalization. A list of primers and thermal protocols is available in supplementary material.

Gene expression

To evaluate the expression of the CX3CR1 gene, RNA was extracted with RNeasy Plus Micro Kit (Qiagen) from cells expanded in vitro for 7–10 days. cDNA was synthetized with Maxima Reverse Transcriptase (Thermo Fisher # EP0742) and CX3CR1 expression was quantified using TaqMan® Assays Hs01922583_s1 and human GAPDH Hs02786624_g1 as housekeeping reference. About 1 ng RNA equivalent was analyzed using the QX200 Droplet Digital PCR System (Bio-Rad). Poisson statistical analysis of the numbers of positive and negative droplets yields absolute quantification of the target sequence.

Flow cytometry

CX3CR1 Monoclonal Antibody (2A9-1) eBioscience™ was used to assess protein expression by flow cytometry in cells expanded in vitro for 7 days. Cells were collected and resuspended in 100 ul of MACS buffer with 2:100 FcR Blocking Reagent (Miltenyi, #130-059-901) and incubated 10’ at 4 °C to avoid aspecific binding of the antibodies. Cells were then incubated with the antibody for 20’ (labeling procedure) at 4 °C. After washing, cells from different tissues were resuspended in MACS buffer (300–400 μl). Vital dye (7AAD) was added, and samples were analysed at BD LSR Fortessa. Results were analysed by FlowJo 10.8.0 software.

Transplantation of edited human hematopoietic stem cells in immunodeficient recipients

Edited human HSPCs were transplanted into 7–8 weeks old NSG females conditioned with busulfan (16.25 mg/kg per 4 days). Cells were administered IV (0.5 × 10⁶/mouse) and ICV (0.3 × 10⁶/mouse) as previously described²⁰. Mice were then provided with syngeneic BMNC for hematopoietic rescue 4 days post-transplant.

Secondary transplantation was performed upon injection of 2 × 10⁶ beads-purified human CD34+ cells (CD34 MicroBead Kit, Miltenyi #130-046-702) harvested from the BM of primary engrafted NSG mice euthanized at 12 weeks post-transplant. Mice were then provided with syngeneic BMNC for hematopoietic rescue 5 days post-transplant. Transplanted mice were monitored by bleeding once/month and euthanized after 12 weeks as previously described.

Evaluation of peripheral engraftment in the blood

Human CD45+ cell engraftment and the presence of edited cells were evaluated by bleeding mice periodically from the tail vein. 0.1 ml of blood was obtained per sample and collected into tubes containing 10 μl EDTA solution (45 ng/ml) to avoid coagulation of the blood during the blood draw. Cells of donor origin were detected by FACS analysis after lysis of erythrocytes with ACK (10 min at room temperature) and specific staining as specified below. The remaining cells were used to extract DNA with QIAamp DNA Micro Kit (QIAGEN) to perform molecular analysis as specified below. A list of flow cytometry antibodies is available in supplementary material.

Mouse tissue collection and processing for flow cytometry and histology

According to the time and the experimental settings, mice were euthanized under deep anesthesia (Ketamine/Xylazine mix) by extensive intracardiac perfusion with cold PBS for 15 min after clumping the femur. Hematopoietic organs and the brain were then collected and differentially processed.

BM cells were harvested by flushing the tibias and femurs with PBS 2%FBS. Spleen and thymus were mechanically disaggregated on a cell strainer (40 μm) in PBS 2%FBS (15 ml for spleen, 5 ml for thymus). About 500 μl of homogenate tissues were centrifuged at 900 × g for 5′ and then resuspended in 100 μl of blocking solution (MACS buffer with 2:100 FcR Blocking Reagent Miltenyi #130-059-901 and 1:100 CD16/CD32 Blocking Assay, BD Biosciences #553142) and incubated 10 min at 4 °C to avoid aspecific binding of antibodies. Cells were then incubated with specific antibodies for 20 min (labeling procedure) at 4 °C. After washing, cells from different tissues were resuspended in MACS buffer (300–400 μl). Vital dye (DAPI or 7AAD) was added, and samples were analysed at BD LSR Fortessa. Results were analysed by FlowJo 10.8.0 software. A list of flow cytometry antibodies is available in supplementary material.

The brain was removed, and the two hemispheres were differently processed. For immunofluorescence analysis, one hemisphere was fixed for 24 h in 4% PFA, embedded in OCT compound, and stored at −80 °C, after equilibration in sucrose gradients (from 10 to 30%) supplemented with 0,02% of sodium azide to avoid contamination. For flow cytometry analysis, cells were obtained by mechanic disaggregation of one brain hemisphere in 1.5 ml in EBSS medium and processed with a papain-based digestion procedure according to the Neural Tissue Dissociation Kit (Miltenyi, #130-092-628). After washing with EBSS medium, the digested suspension was enriched in myeloid cells with Percoll gradient (700×g for 15 min, no break). The cell suspension was then washed with PBS 2%FBS, put in a blocking solution, and stained with specific antibodies as described above for flowcytometric analysis. For the ICV experiment, the brain hemisphere containing the injection site was analysed by FACS analysis, while the contralateral part was used for immunofluorescence analysis.

From all the tissues, a fraction of the cell suspension was stored at −80 °C. DNA was then extracted with the QIAamp DNA Micro Kit (QIAGEN) to perform molecular characterization and evaluate editing efficiency in vivo as described above. A list of primers and thermal protocols is available in supplementary material.

IN-OUT PCR

The frequency of biallelic vs monoallelic editing was assessed by analyzing the in vitro CFU progeny of the infused CD34+ HSPCs and the in vitro CFU progeny of the BM retrieved from the transplanted mice at sacrifice. This assay enables the analysis of clonal genotypes by PCR screening of colonies for targeted integration of a transgene expression cassette³⁰. 800 CD34+ HSPCs and 0.15 × 10⁶ total bone marrow nucleated cells retrieved from transplanted mice at sacrifice were seeded in an enriched methylcellulose-based medium (StemCell Technologies, Cat#H4435) and single-picked after 14 days of in vitro culture. DNA was extracted with DNA Extraction QE Buffer (Biosearch Technologies, Cat# QE09050).

To identify biallelic vs monoallelic editing, we used an “IN-OUT” PCR approach, in which one primer was designed in the targeted genomic locus outside the region of the homology arm (Outside LHA), and the other primer was located inside the transgene cassette (on YFP). A third primer, located in the genomic region on the opposite side of the sgRNA target site from the primer outside the homology arm (on RHA), was also included for identifying alleles without integration. After amplification, DNA was loaded on a 1% Agarose gel, and bands were visualized at ChemiDoc 1708279 (Bio-Rad). A list of primers and thermal protocols is available in supplementary material and methods.

Immunofluorescence analysis

Brains embedded in OCT were serially cut in the sagittal planes on a cryostat in 18-μm sections. Brain sections were obtained from the contralateral side of cell injection for intracerebroventricularly transplanted mice.

Tissue slides were washed twice with PBS, air dried, and blocked with 0.3% Triton and 10% FBS for 1 h at room temperature. Then sections were incubated overnight at 4 °C with primary antibodies diluted in PBS, 0.3% Triton, 10% FBS as follows: rabbit anti-Iba-1 (Wako) 1:250; chicken anti-GFP (Abcam) 1:100; rabbit anti-cherry (Abcam) 1:100; anti-human nuclei (Sigma Aldrich) 1:100. The secondary antibodies goat IgG anti-Chicken Alexa Fluor 488, goat IgG anti-Rabbit Alexa Fluor 488, 546, or 633, goat IgG anti-Rat Alexa Fluor 546 or 633, goat IgG anti-Mouse Alexa Fluor 546 (Molecular Probes, Invitrogen) were diluted 1:500 in PBS, 1% FBS and incubated with sections for 90 min at room temperature. Nuclei were stained with DAPI (Roche) at 1:30 in PBS. Slices were washed in PBS, air dried, and mounted with Fluorsafe Reagent (Calbiochem). Not transplanted mice were used as negative controls for the reporter transgene staining. Incubation with a secondary antibody alone was performed to exclude the background signal. Samples were analyzed with a confocal microscope (Zeiss and Leica TCS SP2; Leica Microsystems Radiance 2100; Bio-Rad; Leica SPE confocal) (λ excitation = 488, 586, 660).

Fluorescent signal was processed by Lasersharp 2000 software. Images were imported into ImageJ software and processed by using automated level correction. For the reconstruction of brain sections, we used a fluorescence microscope Delta Vision Olympus Ix70 for the acquisition of the images, which were then processed by the Soft Work 3.5.0 software. Images were then imported into the Adobe Photoshop CS 8.0 software and reconstructed.

For branching analysis, a Macro was developed in the lab with ImageJ software to perform a standardized analysis on multiple individual cells (>50 per group) from different transplanted mice (n = 3/group). Branching results obtained on fluorescent markers were confirmed on Iba-1stained cells, independently from the fluorescent signal reporter.

Automated Sholl analysis²² was performed by applying the same strategy, to study the radial distribution of microglia branches around the cell body. The sum intersections and a number of intersection radii were selected as parameters to characterize donor-derived cells, in terms of complexity and spatial extension of the cell arborizations, respectively.

Cell sorting from hematopoietic organs of transplanted NSG mice

To evaluate the gene editing efficiency of the hematopoietic compartment in vivo, engrafted cells were sorted from the spleen of transplanted NSG mice as previously described⁵⁴. Briefly, spleens were crushed, and the cell suspension was filtered with a 40-μm cell strainer with cold MACS buffer. The homogenate was then lysed with ACK lysis buffer. After washing with MACS buffer, cells were stained with the dedicated anti-human antibody cocktail for cell lineage sorting. hCD45, hCD3, hCD19, and hCD13 antibodies were used. Dead cells were marked with 7AAD staining. Lymphoid and myeloid populations were sorted with the BD FACSAria II high-speed cell sorter. After sorting, cells were pelleted and stored at −80 °C to proceed with DNA extraction and ddPCR analysis for HDR quantification as described above.

Single-cell data

Single-cell dataset generation

Single cell RNA-Seq was provided by the Single Cell Core at Harvard Medical School, Boston, MA using the 10X Genomics technology. Briefly, MLC sorted from competitively transplanted mice were isolated, and single-cell suspensions were prepared for each sample. Cells were then encapsulated in droplets containing unique barcodes and reverse transcription reagents, followed by library preparation and sequencing.

We generated two libraries for microglia-like cells – MLCs: one for the Cx3cr1^+/+ cells and one for the Cx3cr1^−/+MLCs. Four animals were competitively transplanted with Cx3cr1^+/+ mCherry⁺ HSPCs plus Cx3cr1^−/+ BFP⁺ HSPCs. At sacrifice, 15 days post-transplant, brains were processed to obtain a single cell suspension, and all the mice were pooled to obtain one sample to be stained and sorted as 7AAD⁻/CD11b⁺/mCherry⁺ and 7AAD^–/CD11b⁺/BFP⁺. A total of 2.435 mCherry⁺ cells and a total of 2.634 BFP⁺ cells were sorted and processed for single-cell barcoding in the 10X Genomics platform.

Data processing and quality control

The raw sequencing data was processed using the Cell Ranger software (version 4.0.0)⁵⁵ to obtain gene expression matrices for each sample. The resulting matrices were then imported into the Seurat package (version 4.0.4)⁵⁶ for quality control and downstream analysis. Cells with a low number of detected genes (<350) and high mitochondrial gene content (>15%) were filtered out. Cells with a total number of reads less than 3500 and more than 35,000 were also removed from the dataset. The filtering thresholds have been set based on the inspection of the data distributions and literature.

We employed the ScType algorithm⁵⁷ for automated cluster annotation across the entire single-cell dataset, setting “Brain” as the reference tissue. ScType effectively assigned putative cell types to each cluster. All the smaller clusters displaying association with various other brain cell types, suggesting potential sample contamination, were excluded from downstream analyses to ensure the integrity of the dataset.

Normalization and scaling

The expression data were normalized and scaled using the SCTransform⁵⁸ function in Seurat. This method applies a regularized negative binomial regression to model the count data and correct for technical noise and batch effects. In the normalization process we included as covariates quantitative indexes reflecting cell cycle status (S.Score, G2M.Score), and percentage of cumulative mitochondrial gene expression. Gene Cx3cr1 was removed from the set of genes to avoid biases related to haploinsufficiency. The resulting scaled data was used for all downstream analyses.

Principal component analysis

Principal component analysis (PCA) was performed on the scaled data using the RunPCA function in Seurat. The top 12 principal components were selected based on their contribution to the total variance in the data. The resulting principal components were used as input for UMAP dimensionality reduction.

Uniform manifold approximation and projection (UMAP)

UMAP (McInnes et al. preprint⁵⁹), was performed on the PCA-reduced data using the RunUMAP function in Seurat. This method projects high-dimensional data into a low-dimensional space while preserving the global structure of the data. The resulting UMAP plot was used for visualization and cell type identification.

Clusters identification

Based on the distance matrix calculated using PCA scores, the k-nearest neighbors graph (k = 20, Seurat default) has been calculated. Clusters were identified using the Louvain algorithm. We tested different resolution values (0.6; 1.2) and finally set it at 0.8 based on the robustness of the clustering generated.

Cell type identification

We identified cell cluster marker genes with the FindAllMarkers function in Seurat, which performs differential expression analysis between each cell cluster and the remaining cells in the dataset. We tested SCT normalized data using the t-test option and setting the minimum fraction (min.pct) to 0.1.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

• SEO Powered Content & PR Distribution. Get Amplified Today.
• PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
• PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
• PlatoESG. Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
• PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
• Source: https://www.nature.com/articles/s41467-024-54515-4