Animals

Female C57BL/6 (SM-001), BALB/c (SM-003), C3H/He (SM-008) and NOD-PrkdcscidIl2rgem1/Smoc (M-NSG) mice were purchased from Shanghai Model Organisms Center Inc, China. All mice were housed in groups of 5 individuals per cage and maintained on a 12-h light-dark cycle at 22–25 °C under specific-pathogen free (SPF) conditions. All animal experiments were approved by the Laboratory Animal Research Center, Tongji University. All procedures involving animals were carried out in compliance with the Guide for the Care and Use of Laboratory Animals, and ethical approval was granted by the Ethics Committee, Tongji University. BALB/c mice were thymus injected with AAVs at 7 weeks of age and received T cell depletion at 8 weeks of age. Skin tissue transplantation or pluripotent stem cell transplantation was performed at 4 months of age, 2 months after T cell depletion when the entire T cell repertoire has been re-constituted. The investigators were blinded to allocations during experiments and outcome assessment.

Thymus vaccination

For donor MHC expression in recipient thymus, AAV packaging system was used. In brief, donor MHC cassettes driven by the CMV promoter were constructed in the AAV2/8 vector (Supplementary information, Fig. S1). After packaging, viruses were concentrated through gradient centrifugation and viral titer was detected by qRT-PCR (for rAAV genome).

Mice were anesthetized through intraperitoneal (i.p.) injection of Avertin. Hair on the chest was removed with depilatory cream. Mice were then intubated and connected to a small animal ventilator (RWD, cat. no. R420). After skin disinfection with povidone-iodine, a central skin incision at the level of 2nd intercostal space was made. To expose the thymus, a horizontal incision at the mouse sternum was introduced and set apart with a retractor. AAVs for Ctrl TV or DMTV (1 × 1013 vg/mL, 10 μL for each MHC expression virus, and the same total dosage was applied for Ctrl in each group) were intrathymically injected with a 30-gauge Hamilton syringe. During the thymus vaccination procedure, mice were maintained inflated with a ventilator before thoracic cavity was closed and the opening was closely sutured. Carprofen (5 mg/kg, subcutaneous injection) and enrofloxacin (10 mg/kg, i.p. injection) were used to provide analgesia or prevent infection for 3 days after surgery.

T cell depletion

1.5 mg anti-CD4 (BioXCell, New Hampshire, USA, BP0003-1) and 0.8 mg anti-CD8 (BioXCell, BP0061) monoclonal antibodies (mAbs) were i.p. injected into Ctrl TV mice or DMTV mice twice (7 days and 10 days after thymus vaccination) to deplete pre-existing CD4+ and CD8+ T cells, respectively. On day 10 after thymus vaccination, mice were subjected to a 3 Gy TBI. The populations of CD4+ and CD8+ T cells were then assessed by flow cytometry analysis.

TEC, DC and PBMC isolation

To validate ectopic MHC expression in TECs and DCs, vaccinated thymus was isolated and cut into small pieces with scissors. Thymus tissues were then digested with 0.5 mg/mL papain (Sangon, Shanghai, China, cat. no. A003124), 2.5 mg/mL collagenase IV (R&D, cat. no. 9001-12-1) and 0.1 mg/mL DNase I (Sigma-Aldrich, Darmstadt, Germany, cat. no. 11284932001) in basal DMEM/F12 medium at 37 °C for 30 min. 10% fetal bovine serum (FBS) in DMEM/F12 was applied to stop the digestion and cells were passed through a 70-μm cell strainer before flow cytometry analysis.

PBMCs were collected for MLR assay, DSA detection, single-cell sequencing and immune cell composition detection. Peripheral blood cells and spleen homogenates (passed through a 40-μm cell strainer) were collected in tubes prefilled with EDTA. Histopaque®-1083 (Sigma-Aldrich, cat. no. 10831-100 mL) and Histopaque®-1077 (Sigma-Aldrich, cat. no. 10771-100 mL) were then used to enrich mouse and human PBMCs via density gradient centrifugation, respectively.

Flow cytometry analysis

Flow cytometry analyses were performed as previously described. Cells were washed with flow cytometry buffer (PBS containing 2% BSA and 2 mM EDTA) and collected by centrifugation at 400× g for 5 min. Cells resuspended in flow cytometry buffer were then incubated with anti-mouse CD16/CD32 mAb (BD, New Jersey, USA, 553142) for 10 min on ice to block nonspecific FcR binding, followed by incubation of fluorescently-labeled antibodies for 30 min on ice. Flow cytometry was performed on a FACSVerse™ flow cytometer (BD). FlowJo Software was used for data analysis.

Antibodies used in flow cytometry analyses are as follows, CD45 (eBioscience, California, USA, cat. no. 12-0451-82; isotype: rat IgG2b kappa), CD45 (eBioscience, cat. no. 17-0451; isotype: rat IgG2b kappa), CD326 (BD, cat. no. 563478; isotype: rat IgG2a κ), CD11c (eBioscience, cat. no. 12-0114-81; isotype: Armenian hamster IgG), H2-Kb (eBioscience, cat. no. 11-5958-82; isotype: mouse IgG2a kappa), H2-Db (BD, cat. no. 553573; isotype: mouse IgG2b, κ), IA-b (Biolegend, San Diego, CA, cat. no. 116405; isotype: mouse IgG2a, κ), H2-Kd/Dd (Biolegend, cat. no. 34-1-2S; isotype: mouse IgG2a), CD3 (eBioscience, cat. no. 11-0031-63; isotype: Armenian hamster IgG), CD4 (BD, cat. no. 553051; isotype: rat IgG2a κ), CD8 (eBioscience, cat. no. 12-0081-82; isotype: rat IgG2a, κ), huCD45 (BD, cat. no. 555485; isotype: mouse IgG1, κ) and huCD3 (eBioscience, cat. no. 11-0038-42; isotype: mouse IgG1, κ).

MLR assay

Isolated PBMCs were collected in X-VIVO™ 15 medium (Lonza, Visp, Switzerland, cat. no. 04-418Q) supplemented with 10% FBS and 200 U/mL pen/strep (Gibco, Massachusetts, USA, cat. no. 15140122). 1.0 × 105 PBMCs from either Ctrl TV BALB/c or DMTVC57 BALB/c mice were labeled with CellTrace™ CFSE (Thermo Fisher Scientific, Massachusetts, USA, cat. no. C34554) and incubated with irradiated (25 Gy) 1.0 × 105 PBMCs from BALB/c, C57BL/6 or C3H/He mice for 10 days in a well of 96-well flat-bottom culture plate in culture medium (X-VIVO™ 15, 10% FBS, 2mM L-glutamine (Gibco, cat. no. A2916801), 50 μM β-mercaptoethanol (Sigma-Aldrich, cat. no. 60-24-2), 20 U/mL IL-2 (Biolegend, cat. no. 575402) and 200 U/mL pen/strep. Phytohemagglutinin-L (PHA-L) (Invitrogen, Massachusetts, USA, cat. no. 00-4977-93) was served as a positive control. CFSE intensity in CD3+ T cells was checked on a FACSVerse flow cytometer (BD) and data were processed with FlowJo.

Integrated scRNA-seq & scTCR-seq

10× Genomics platform was used for integrated scRNA-seq & scTCR-seq according to the manufacturer’s protocols. Libraries were sequenced by Illumina sequencer (Illumina, San Diego, CA) on a 150 bp paired-end run.

Cellranger (v7.1.0) was used to align reads to mm10 genome and generate feature-barcode matrices. Genes expressed in fewer than 3 cells were filtered from expression matrices. Cells with a mitochondrial fraction not in the highest confidence interval were filtered out, which results in removal of cells with a mitochondrial percentage of more than 5%.

Doublets were excluded with DoubletFinder (v2.0.3). Artificial doublets were generated from raw RNA count matrices based on the average of gene expression profiles of randomly sampled cell pairs. After merging artificial doublets with real existing scRNA-seq data using Seurat (v4.0.4), euclidean distance matrix was obtained from cell embeddings in principal component (PC) spaces. The proportion of artificial nearest neighbors (pANN) is computed by dividing its number of artificial neighbors by the neighborhood size (pK). Cells with the highest pANN were identified as doublets. The parameters were set as follows, where doublets proportion pN = 0.25, PCs = 30, pK was determined using mean-variance-normalized bimodality coefficient.

Seurat package (v4.0.4) was used for clustering. Raw RNA count matrices were normalized using SCTransform function with mitochondrial fraction as a variable to regress out. Top 2000 features that are repeatedly variable across datasets for integration were identified with SelectIntegration function. Anchors were then determined using the FindIntegrationAnchors() function and datasets were integrated together with IntegrateData() function. Dimensionality reduction was performed on the integrated data with principal component analysis (PCA). First 20 PCs were then used further for UMAP visualization and clustering procedure. The resolution of 0.1 was used in FindClusters function after computing the nearest neighbors by FindNeighbors function. Differentially expressed genes (DEGs) between clusters were identified using FindAllMarkers function. Cell type annotation was carried out with the expression of canonical gene markers.

Single-cell VDJ receptor sequences were assembled and analyzed with Cell Ranger’s vdj pipeline (v7.1.0). T cells with inappropriate combinations of α- and β-chains were removed. The expression levels of specific TCRs were assigned to cell populations defined with scRNA-seq clustering and visualized with UMAP.

Monitoring DSAs

Monitoring DSAs was used for analyzing the overall immunoreactivity in DMTV-treated mice upon donor cell priming. Irradiated PBMCs from BALB/c, C57BL/6 and C3H/He mice were tail vein injected into the Ctrl TV BALB/c mice and DMTVC57 BALB/c mice. 10 days after inoculation, their sera were collected and incubated with the corresponding irradiated PBMCs at 37 °C overnight. FITC-anti-mouse secondary antibody (JacksonImmuno, Pennsylvania, USA, cat. no. 115-095-003) incubation was performed on the next day for 1 h at room temperature followed by flow cytometry analysis.

Skin transplantation

For skin transplantation, mice were anaesthetized with 4% isoflurane (RWD, Shenzhen, China, cat. no. R510-22-10) in medical air and maintained under anesthesia using a nose cone with 1.5% isoflurane. Animals were placed on a heat pad set at 37 °C and hair was trimmed from the back of both recipient and donor. A 9 mm × 9 mm piece of full thickness skin was then cut off from the recipient back, and a 10 mm × 10 mm full thickness skin collected from the back of a donor was laid smoothly to align with the edge of the cut skin and was subsequently sutured together. Recipients after transplantation were patched with 3M nexcare (3M, Kleinostheim, Germany, cat. no. CBGBLRUS1509) and 3M athletic wrap (3M, cat. no. CBGBLRUS1507) was used for secondary fixation to make the skin better fit into the recipient graft bed. Carprofen and enrofloxacin were used to provide analgesia or prevent infection.

Staining

Skin grafts and surrounding tissues or recovered teratomas were collected and fixed in 4% paraformaldehyde (PFA) at 4 °C overnight followed by gradient sucrose (in PBS) treatment. Tissues were then embedded in OCT compound (Sakura, California, USA, cat. no. 4583) and sectioned at 10-μm thickness using LEICA CM3050 S. For immunofluorescence staining, slides were incubated in blocking buffer (10% donkey serum, 0.1% Triton X-100 in PBS) for 1 h and incubated with primary antibodies at 4 °C overnight. After adequate washing with PBS, slides were incubated with fluorescently-conjugated secondary antibodies for 1 h at room temperature. Nuclei were counterstained with Hoechst 33258 (Sigma-Aldrich, D9542). Slides were then mounted with Fluoromount-G Mounting Medium (Southern Biotech, Alabama, USA, cat. no. 0100-01). Images were captured using confocal microscope (Leica SP8).

For hematoxylin and eosin (H&E) staining, tissues were fixed in 4% PFA and subjected to paraffin embedding. Paraffin-embedded tissues were sectioned at 3-μm thickness, and sections were processed for H&E staining.

The following primary antibodies were used for staining analyses in the current study, anti-CD3 (Servicebio, Wuhan, China, GB13014), anti-CD4 (Servicebio, GB13064-2) and anti-CD8 (Servicebio, GB114196).

Construction of humanized mice

BLT-humanized mouse model was constructed according to a previously published protocol.46 Normal aborted fetuses were obtained from Shanghai First Maternity and Infant Hospital or Jing’an District Hospital of Traditional Chinese Medicine with agreement of the donors and approval of related ethical review and informed consent documents. All the procedures were approved by the Ethics Committee of School of Medicine, Tongji University, and complied with the fundamental guidelines for the proper conduct of Interim Measures for the Administration of Human Genetic Resources and related activities in academic research institutions under the jurisdiction of the Chinese Ministry of Health.58 HLA typing (HLA-A, -B and -Cw) of human fetus was performed by the Shanghai Tissuebank Biotechnology Co., Ltd. M-NSG mice were exposed to 1 Gy irradiation right before renal subcapsular transplantation of fetal liver and thymus tissues at the right side. The thymus/liver-transplanted M-NSG mice were then tail vein injected with 5.0 × 105 huCD34+ HSCs enriched with magnetic beads (Meltenyi, Bergisch Gladbach, Germany, 130-056-701) from the liver of the same fetal donor.

Teratoma formation

mESCs or hESCs were injected subcutaneously over the scapula in recipient mice as indicated at a dosage of 1.0 × 106 cells for mESCs and 2.0 × 106 cells for hESCs per injection site.

Statistical analysis

Statistical analyses were carried out with GraphPad Prism 9. Differences among groups were evaluated by Student’s t-test, One-way ANOVA or Log-rank test as indicated in each figure legend. One-way ANOVA of variance with a Dunnett’s multiple comparisons test was performed to identify statistical significance among groups of samples. All data are presented as mean ± SEM. Significance was determined with ****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05.