Mice

C57BL/6J (RRID: IMSR_JAX: 000664), NSG (NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ, RRID: IMSR_JAX: 005557)8, NBSGW (NOD.Cg-KitW-41JTyr+PrkdcscidIl2rgtm1Wjl/ThomJ, RRID: IMSR_JAX:026622)12, NSGW41 (NOD.Cg-KitW-41JPrkdcscidIl2rgtm1Wjl/WaskJ, RRID: IMSR_JAX:026497)11 and JAX NSG huCD34 (RRID: IMSR_JAX: 005557) mice (Supplementary Table 8) were purchased from The Jackson Laboratory (JAX NSG huCD34 mice were constructed by grafting γ-irradiated 3-week-old female NSG mice with cord blood huCD34+ HSCs). In all experiments, male and female mice were used in virtually equal proportions.

huCD34+ HSCs to be used for construction of THX, huNBSGW, huNSGW41 and huNSG mice were isolated from human umbilical cord blood collected immediately after cesarean section from full-term, normally developed male and female newborns (in approximately equal numbers) upon informed consent from healthy puerperae (18–45 years old with no infectious disease or history of cancer) of different ages, races and ethnic backgrounds (Supplementary Table 9; Department of Obstetrics and Gynecology, The University of Texas Long School of Medicine). CD34+ cells were purified using EasySep Human CD34 Positive Selection Kit II (17856, STEMCELL Technologies) according to the manufacturer’s instructions, yielding at least 99% huCD34+ cell preparations. Freshly purified huCD34+ cells were resuspended in PBS supplemented with 2% FBS for immediate grafting or frozen in 10% dimethylsulfoxide, 72% FBS, 18% RPMI medium and kept in liquid nitrogen for later grafting.

huNSG mice (Supplementary Table 10) were constructed by myeloablative conditioning of NSG mice neonates (within 48 h of birth) with (1 Gy) γ-radiation, followed by intracardiac (left ventricle) injection with purified cord blood huCD34+ cells (1.5 × 105 freshly isolated or frozen-thawed huCD34+cells in 50 μl PBS supplemented with 2.0% FBS) using a 27-gauge needle. huNBSGW (Supplementary Table 11) and huNSGW41 mice were constructed by grafting non-γ-irradiated, genetically myeloablated NBSGW and NSGW41 mice neonates (within 48 h of birth) intracardially (left ventricle) with cord blood huCD34+ cells. THX mice were generated by feeding huNBSGW or huNSGW41 mice E2 (3301, Sigma-Aldrich; 1.5 μM in drinking water resulting in a dose of 6.1 × 10−4 mg per kg body weight per day) ad libitum starting at 14–18 weeks of age (18 weeks in most cases) and continuing thereafter. After 4 weeks of E2 conditioning, huNBSGW or NSGW41 mice (referred to as THX mice; Supplementary Table 12) were ready for experiments or continued E2 for later use. E2 conditioning of huNBSGW or huNSGW41 mice did not start before 14 weeks of age, as estrogen (albeit at high dose) might inhibit early thymus development by T cell proliferation.

Most THX mice were constructed using NBSGW mice as only a dozen NSGW41 mice were acquired in 2019 from The Jackson Laboratory before the sale of such mice was discontinued. NSGW41-based THX mice were used in the human antibody response to NP-CGG (n = 3), in the lupus studies as part of the healthy THX controls (n = 4 of 12) as well as for generation of Lupus THX mice (n = 5) as described in ‘Lupus THX mice, human autoantibodies, immunopathology and mortality’. huCD45+ cells in blood, spleen and BM of humanized mice were identified by flow cytometry using APC-anti-huCD45 monoclonal antibody (304011, BioLegend; 1:100 dilution) and Pacific Blue-anti-moCD45 monoclonal antibody (103125, BioLegend; 1:100 dilution). Generally, THX and huNBSGW mice displayed up to 96.1% and 89.3% huCD45+ cells in circulating blood, respectively. huNSG and JAX NSG huCD34 mice displayed, at peak, approximately 45% and 20% huCD45+ cells, respectively. Circulating huCD45+ mononuclear cell numbers (cells per ml of blood) were measured by complete blood count analysis, in which blood was collected in EDTA-coated microtubes and analyzed using a XT2000iV or XE-5000 blood analyzer (Sysmex). THX, huNBSGW and huNSG mice used in all experiments were 20 to 24 weeks of age, unless indicated otherwise. JAX NSG huCD34 mice were 23 weeks of age. Mice used in all experiments were housed in a pathogen-free barrier animal vivarium facility at The University of Texas Health Science Center at San Antonio and were free of infection or disease. Housing rooms were maintained at a 14-h light/10-h dark cycle and controlled temperature of approximately 22–23 °C with 40–60% humidity. Food (Teklad LM-485 Sterilizable Mouse/Rat Diet, 7912, Inotiv) and water were sterilized.

Estrogen

Serum estradiol concentrations in non-intentionally immunized THX and huNBSGW mice (18–24 weeks old) were measured using Cayman Estradiol ELISA Kit (501890, Cayman Chemical), according to the manufacturer’s instructions, and compared to mice and human physiological range80,81,82,83,84,85,86. This platform uses an estradiol acetylcholinesterase conjugate (estradiol acetylcholinesterase Tracer) in an inhibition/competition assay, measuring serum estradiol concentration by OD at 414 nm. High OD readings reflect low estradiol concentrations, while low OD readings reflect high concentrations. Sera were collected from equal numbers of male and female mice, with female mice sera collected generally during proestrus, metestrus and diestrus.

FACS and CyTOF

For the cell surface FACS analysis, cells from blood of healthy humans or blood, BM (tibia and femur), thymus, spleen, LNs (cervical, mediastinal, axillary, mesenteric) and/or Peyer’s patches of humanized mice (THX, huNBSGW, huNSG or JAX NSG huCD34 mice) were surface stained with fluorochrome-conjugated monoclonal antibodies (Supplementary Table 13) in Hank’s Buffered Salt Solution (HBSS, MT21022CM, Fisher Scientific) plus 0.1% bovine serum albumin (BSA, BP1600-100, Fisher Scientific; BSA-HBSS) for 20 min. After washing, cells were resuspended in BSA-HBSS for flow cytometry. In vitro-stimulated and/or ex vivo mononuclear cells were stained with FITC-anti-huCD45 monoclonal antibody (clone 30-F11, 368507, BioLegend; 1:100 dilution), PE-anti-huCD45 monoclonal antibody (clone 2D1, 368509, BioLegend; 1:100 dilution), Pacific Blue-anti-moCD45 monoclonal antibody (clone 2D1, 103125, BioLegend; 1:100 dilution), PE-anti-huCD19 monoclonal antibody (clone HIB19, 302208, BioLegend; 1:100 dilution), PE-Cyanine7-anti-huCD19 monoclonal antibody (clone HIB19, 302216, BioLegend; 1:100 dilution), FITC-anti-huCD20 monoclonal antibody (clone 2H7, 302303, BioLegend; 1:100 dilution), BV510-anti-huCD138 monoclonal antibody (clone MI15, 356517, BioLegend; 1:100 dilution), PE-anti-huIgM monoclonal antibody (clone MHM-88, 314507, BioLegend; 1:100 dilution), BV510-anti-huIgM monoclonal antibody (clone MHM-88, 314521, BioLegend; 1:100 dilution), BV650-anti-huIgM monoclonal antibody (clone MHM-88, 314525, BioLegend; 1:100 dilution), APC-Fire 750-anti-huIgM monoclonal antibody (clone MHM-88, 314545, BioLegend; 1:100 dilution), BV421-anti-huIgD monoclonal antibody (clone IA6-2, 348225, BioLegend; 1:100 dilution), BV785-anti-huIgD monoclonal antibody (clone IA6-2, 348241, BioLegend; 1:100 dilution), BV421-anti-huIgG monoclonal antibody (clone M1310G05, 410703, BioLegend; 1:100 dilution), FITC-anti-huIgG monoclonal antibody (clone G18-145, 555786, BD Pharmingen; 1:100 dilution), APC-anti-huIgA monoclonal antibody (clone IS11-8E10, 130-113-427, Miltenyi Biotec; 1:50 dilution), FITC-anti-huIgA (c31577, Invitrogen; 1:100 dilution), APC-Fire 750-anti-huIgE monoclonal antibody (clone MHE-18, 325515, BioLegend; 1:100 dilution), APC-Cyanine7-anti-huCD11c monoclonal antibody (clone Bu15, 337217, BioLegend; 1:100 dilution), APC-anti-huCD14 monoclonal antibody (clone 63D3, 367117, BioLegend; 1:100 dilution), BV786-anti-huCD56 monoclonal antibody (clone 5.1H11, 362549, Biolegend; 1:100 dilution), PE-anti-huCD27 monoclonal antibody (clone M-T271, 356405, BioLegend; 1:100 dilution), BV650-anti-huCD38 monoclonal antibody (clone HB-7, 356619, BioLegend; 1:100 dilution), PE-Cyanine7-anti-huCD5 monoclonal antibody (clone UCHT2, 300621, BioLegend; 1:100 dilution), Super Bright 600-anti-huCD3 monoclonal antibody (clone OKT3, 63003741, eBioscience; 1:100 dilution), APC-anti-huCD4 monoclonal antibody (clone A161A1, 357407, BioLegend; 1:100 dilution), BV421-anti-huCD4 monoclonal antibody (clone A161A1, 357423, BioLegend; 1:100 dilution), PE-anti-huCD8 monoclonal antibody (clone SK1, 344705, BioLegend; 1:100 dilution), Alexa Fluor 700-anti-huCD8 monoclonal antibody (clone SK1, 344723, BioLegend; 1:100 dilution), PE-anti-huCXCR5 monoclonal antibody (clone J252D4, 356903, BioLegend; 1:100 dilution), FITC-anti-huCXCR5 monoclonal antibody (clone J252D4, 356913, BioLegend; 1:100 dilution), FITC-anti-huPD-1 monoclonal antibody (clone NAT105, 367411, BioLegend; 1:100 dilution), PE-Cyanine7-anti-huPD-1 monoclonal antibody (clone A17188B, 621615, BioLegend; 1:100 dilution), Pacific Blue-anti-huICOS monoclonal antibody (clone C398.4A, 313521, BioLegend; 1:100 dilution), PE-anti-huEpCAM (clone EPR20532-225, ab237397, Abcam; 1:100 dilution), PE-Cyanine7-anti-moEpCAM (clone G8.8, 118216, BioLegend; 1:100 dilution), APC-anti-huHLA-A,B,C (MHC I) monoclonal antibody (clone W6/32, 311409, BioLegend; 1:100 dilution), FITC-anti-huHLA-DR, DP, DQ (MHC II) monoclonal antibody (clone Tü39, 361705, BioLegend; 1:100 dilution) or 7-AAD (A9400, Sigma-Aldrich). For analysis of human red blood cells and platelets, THX mice red blood cells were stained with APC-anti-moTER-119 monoclonal antibody (clone TER-119, 116211, BioLegend; 1:100 dilution) and FITC-anti-huCD235a monoclonal antibody (clone HI264, 349103, BioLegend; 1:100 dilution). THX mice (low forward scatter) platelets were stained with PE-Cyanine7-anti-moCD41 monoclonal antibody (clone MWReg30, 133915, BioLegend; 1:100 dilution) and PerCp-anti-huCD61 monoclonal antibody (clone VI-PL2, 336409, BioLegend; 1:100 dilution).

For the intracellular FACS analysis, AID-expressing and BLIMP-1-expressing huB cells and huPBs/PCs (2.0 × 106 cells) were surface stained with anti-huCD45, anti-huCD19, anti-huCD27, anti-huCD38 and anti-huCD138 monoclonal antibodies, as well as Fixable Viability Dye eFluor 780 (65-0865-14, Fisher Scientific). After washing, cells were fixed in Cytofix/Cytoperm buffer (554655, BD Biosciences, 250 µl) and incubated at 4 °C for 45 min. Cells were washed twice in BD Perm/Wash buffer (554723, BD Biosciences) for permeabilization and stained with Alexa Fluor 647–anti-huAID antibody (bs-7855R-FITC, Bioss; 1:50 dilution) or Alexa Fluor 488-anti-huBLIMP-1 monoclonal antibody (clone 646702, IC36081G, R&D Systems; 1:50 dilution) in BD Perm/Wash buffer for 30 min at 4 °C. Cells were washed again twice in BD Perm/Wash buffer and resuspended in BSA-HBSS for flow cytometry.

Flow cytometry analysis and sorting were performed using single-cell suspensions. Cells were gated by forward and side scattering to exclude dead cells and debris (Supplementary Fig. 1a–c). Cell analysis was performed on pre-gated huCD45+ cells using a BD LSR-II or FACS Celesta flow cytometer (BD Biosciences) with FACSDiva software v9.4 (BD Biosciences). Data were acquired and analyzed using FlowJo v10.9 (Tree Star).

To assess human immune lymphoid and myeloid cell reconstitution in THX mice, single-cell suspensions of splenic white cells from non-intentionally immunized THX mice (20–24 weeks old) were incubated for 30 min at 4 °C with a 50 μl cocktail of metal conjugated anti-human monoclonal antibodies (Supplementary Table 2) from the MaxPar Direct Immune Profiling Assay, 30 Marker Kit (201325, Fluidigm), followed by washing for 10 min at room temperature. Cell viability was measured by DNA intercalation (Cell-ID Intercalator-103Rh). Labeled cells were analyzed by Helios mass cytometer (CyTOF software v6.7, Fluidigm) using a flow rate of 0.045 ml min−1. Human immune lymphoid and myeloid cell population frequencies, quality-control metrics and data plot displays were acquired using Maxpar Pathsetter software v3.0 (401018, Fluidigm).

Bacteria-bound huIgD and huIgA in THX and huNBSGW mice were detected as we described38,78. Briefly, feces (10 mg) were suspended in 100 μl PBS, homogenized and centrifuged at 400g for 5 min to remove large particles. Supernatant was centrifuged at 8,000g for 10 min, then analyzed for free huIgD and huIgA by ELISA. To detect bacteria-bound huIgD and huIgA, the bacterial pellet was resuspended in 1 ml PBS containing 1.0% (wt/vol) BSA. After fixation with 7.2% formaldehyde for 10 min at room temperature, bacteria were washed with PBS, stained with FITC-anti-huIgD (clone IA6-2, 348205, BioLegend; 1:100 dilution) or APC-anti-huIgA (clone IS11-8E10, 130-113-427, Miltenyi Biotec; 1:50 dilution) monoclonal antibodies on ice for 30 min, washed again then resuspended in PBS containing 0.2 μg ml−1 DAPI for flow cytometry analysis. All events revealed by DAPI were considered as bacteria.

Human mononuclear cells

huPBMCs were isolated from buffy coats obtained from healthy male and female human donors of different ages (18–65 years old), races and ethnic backgrounds (South Texas Blood and Tissue Center; Supplementary Table 14). The buffy coat was diluted at a 1:2 ratio in sterile PBS (pH 7.4, BP3991, Fisher Scientific) and then applied to a Histopaque-1077 density gradient (10771, Sigma-Aldrich) in 50 ml SepMate tubes (85450, STEMCELL Technologies), which were spun at 1,000g for 10 min. Recovered huPBMCs were washed in PBS and resuspended in RPMI (10-040-CV, Corning RPMI-1640 medium) supplemented with 10% vol/vol Hyclone FBS (42Q7980K, Gibco) and 1% vol/vol antibiotic-antimycotic solution (penicillin–streptomycin and amphotericin B, SV30079.01, Cytiva Life Sciences; FBS-RPMI).

Human immune cells were isolated from humanized mouse blood, BM, thymus, spleen, LNs and/or Peyer’s patches, and suspended in ACK Lysis Buffer (BP10-548E, Lonza) to lyse erythrocytes. Peripheral blood (approximately 250 μl) was collected from the submandibular vein into microtubes containing heparin (H19, Fisher Scientific; 25 μl, 1,000 units per ml). After quenching with FBS-RPMI and centrifugation, erythrocyte-free cells were resuspended in FBS-RPMI for further preparation or analysis.

Differentiation of naive human B cells from humans and from THX mice

To analyze CSR, PC and MBC differentiation, naive huCD19+IgM+IgD+B cells were isolated from huPBMCs obtained from healthy participants by negative selection using EasySep Human Naive B Cell Isolation Kit (17254, STEMCELL Technologies), according to the manufacturer’s instructions, yielding at least 98% huCD19+IgM+IgD+B cells. After pelleting, huB cells were resuspended in FBS-RPMI before further analysis or stimulation. Naive huCD19+IgM+IgD+B cells were isolated from THX mouse spleens by negative selection using biotin-anti-huCD43 (9620-08, clone DF-T1, SouthernBiotech; 1:50 dilution) and biotin-anti-huCD3 monoclonal antibodies (300403, clone UCHT1, BioLegend; 1:50 dilution) followed by positive selection using biotin-anti-huIgD monoclonal antibody (348212, clone IA6-2, BioLegend; Supplementary Table 15) and MagniSort Streptavidin Positive Selection Beads (MSPB-6003-74, Thermo Fisher Scientific), yielding at least 98% huCD19+IgM+IgD+B cells. After pelleting, B cells were resuspended in FBS-RPMI. Naive huIgM+IgD+B cells from humans or THX mice were cultured in FBS-RPMI (5.0 × 105 cells per ml) for 72 h (for RNA transcript analysis) or up to 120 h (for flow cytometry analysis) upon stimulation with: membrane-CD154 (3.0 U ml−1)38,39,78 or CpG ODN 2395 (Eurofins Genomics, 2.5 μg ml−1) plus recombinant huIL-2 (589102, BioLegend, 100 ng ml−1), recombinant huIL-4 (574002, BioLegend, 20 ng ml−1) and recombinant huIL-21 (571202, BioLegend, 50 ng ml−1) for CSR to huIgG. For CSR to huIgA, naive huIgM+IgD+ B cells were cultured under similar conditions upon stimulation with membrane-CD154 or CpG ODN 2395 plus recombinant huIL-2, recombinant huIL-21, recombinant TGF-β (781802, BioLegend, 4.0 ng ml−1) and recombinant retinoic acid (11017, Cayman Chemicals, 4.0 ng ml−1). Pre-gated huCD45+huCD19+ cells were stained with specific human monoclonal antibodies (Supplementary Table 13) to detect huIgM+, huIgD+, huIgG+, huIgA+ or huIgE+ B cells, huCD27+CD38+ PBs and class-switched huCD27+IgD− memory-like B cells by flow cytometry.

huBCR IgM+ B cell and huTCR repertoires and huIgM+ B and T cell clonality

To analyze expressed huVHDJH-Cμ, huVκJκ-Cκ and huVλJλ-Cλ or huVαJα-Cα and huVβJβ-Cβ gene repertoires, huIgM+ B cells and huT cells were isolated from blood of healthy humans (Supplementary Table 14) and spleens of non-intentionally immunized THX mice (20–24 weeks old). RNA (2 µg) was extracted using RNeasy Mini Kit (74104, Qiagen). huVHDJH-Cμ, huVκJκ-Cκ and huVλJλ-Cλ or huVαJα-Cα and huVβJβ-Cβ mRNA transcripts were reverse transcribed from huIgM+ B or huT cell RNA by RT-5′ RACE PCR using SuperScript III First-Strand Synthesis System (18080051, Invitrogen) and a huCμ-, huCκ-, huCλ-, huCα- or huCβ-specific reverse primer (Supplementary Table 16). Single-strand cDNA was cleaned up using QIAquick PCR purification kit (28104, QIAGEN) and 3′ poly-dA tailed by TdT and dATP. The dA-tailed cDNA was then amplified by PCR using a forward oligo-dT primer together with a nested huCμ-, huCκ-, huCλ-, huCα- or huCβ-specific reverse primer. Both forward and reverse primers were tagged with Illumina overhang adaptors. PCR amplification conditions were 95 °C for 30 s, 55 °C for 30 s and 72 °C for 40 s for 35 cycles. cDNA amplicons were cleaned up using QIAquick PCR purification kit, further amplified by index PCR involving Illumina clustering adaptors and beads cleanup, quantified and then loaded onto the Illumina MiSeq system using the 300-bp pair-end sequencing module. huVHDJH-Cμ, huVκJκ-Cκ, huVλJλ-Cλ, huVαJα-Cα and huVβJβ-Cβ repertoires were analyzed using IMGT/HighV-QUEST v1.9.2 (The International ImMunoGeneTics Information System; http://www.imgt.org/HighV-QUEST/home.action/).

To identify individual huB and huT cell clones and analyze huB or huT cell clonal diversity, huB cell VHDJH-Cµ or huT cell VβDJβ-Cβ transcripts (up to 250,000 sequences) of healthy humans and THX mice were analyzed by Illumina MiSeq amplicon sequencing and segregated based on the same huVH or huVβ gene segment, the same and unique huIgH or huTCRβ CDR3 together with the same huJH or huVβ sequence87,88,89,90,91. Each discrete clone was depicted as an individual rectangle or square (unique color), whose area reflects huB or huT cell clone size, as inferred from the sum of identical huVHDJH-Cµ or huVβDJβ-Cβ transcripts (TreeMaps, Microsoft Excel v16.83 and IMGT/HighV-QUEST statistic data).

THX mice huB cell SHM/CSR, clonality and intraclonal diversification

To analyze SHM in the NP16-CGG-induced antibody response, RNA (2 µg) was extracted from THX mice total and sorted NP16-specific huB cells using the RNeasy Mini Kit (74104, Qiagen), and cDNA was synthesized using the SuperScript III First-Strand Synthesis System (18080051, Invitrogen) with oligo-dT primer. Rearranged huV1DJH-Cγ, huV3DJH-Cγ, huV1DJH-Cα1 and huV3DJH-Cα1 cDNA was amplified using a huV1 or huV3 leader-specific forward primer together with a nested huCγ- or huCα-specific reverse primer tagged with Illumina overhang adaptors (Supplementary Table 16) and Phusion high-fidelity DNA polymerase (M0530S, New England BioLabs)—amplification of huIgH V1 and V3 genes was chosen as these families include gene members of high sequence similarity to mouse V1-72 (V186.2/V3 gene), the gene encoding the most efficient ‘NP-binding’ mouse IgH V segment (https://www.imgt.org/ligmdb/view?id=J00239/)36,92. PCR amplification conditions were 98 °C for 10 s, 60 °C for 45 s and 72 °C for 1 min for 30 cycles. The cDNA amplicons were further amplified and sequenced as described in ‘huBCR IgM+ B cell and huTCR repertoires and huIgM+ B and T cell clonality’. Somatic point mutations in recombined transcripts were analyzed using IMGT/HighV-QUEST v1.9.2 (https://www.imgt.org/HighV-QUEST/login.action/) and corrected for polymerase and sequencing error rates (0.008) to calculate the frequency of somatic point mutations. To analyze huB cell clonality and SHM in the DNP-CpG-, S. Typhimurium flagellin-, Pfizer COVID-19 mRNA- and RBD–KLH-induced antibody responses, THX mouse huB cell VHDJH-Cμ, VHDJH-Cγ, VHDJH-Cα, VκJκ-Cκ or VλJλ-Cλ transcripts were reverse transcribed, amplified and sequenced as described in ‘huBCR IgM+ B cell and huTCR repertoires and huIgM+ B and T cell clonality’, then analyzed for point mutations as described above.

B cell clonal diversity in immunized THX mice was analyzed as described in ‘huBCR IgM+ B cell and huTCR repertoires and huIgM+ B and T cell clonality’. To analyze intraclonal diversification, shared and unique point mutations in huVHDJH-CH transcripts within each huB cell clone were used to construct genealogical trees (phylogenetic maps), revealing sequential multistep accumulation of point mutations from unmutated progenitors, and allowing for detailed intraclonal diversification analysis. Genealogical trees were constructed by uploading FASTA files of all segregated huVHDJH-CH transcripts onto PHYLOViZ Online v2.0 (http://www.phyloviz.net/), which uses a JAVA implementation of the Feil’s goeBURST algorithm rules for visualization of multiple phylogenetic inference trees.

To quantify AICDA, PRDM1, huVHDJH-Cμ, huVHDJH-Cγ1, huVHDJH-Cα1 and huVHDJH-Cε transcript expression in huB cells from THX mice in vitro and ex vivo and huB cells from humans in vitro, RNA extraction and cDNA synthesis were performed as described above. Transcript expression was analyzed by SYBR Green dye (IQ SYBR Green Supermix, 115010139, Bio-Rad) incorporation in PCR reactions involving specific forward and reverse primers (Supplementary Table 16). Reactions were performed in an iCycler (Bio-Rad) real-time qPCR system under the following amplification cycles: 95 °C for 15 s, 40 cycles at 94 °C for 10 s, 60 °C for 30 s and 72 °C for 30 s—data acquisition was performed during this 72 °C extension step (Bio-Rad CFX Manager Software v3.1). Melting curve analysis was performed from 72 to 95 °C. The 2−ΔCt method (2−ΔCt = 2-[Ct(HPRT1)-Ct(target gene)]) was used to determine levels of transcripts, and data were normalized to levels of human HPRT1.

Humanized mice antibody response to conjugated haptens

THX, huNBSGW (20–24-week-old) and JAX NSG huCD34 (23-week-old) mice were injected i.p. with 4-hydroxy-3-nitrophenylacetyl (NP) conjugated to chicken gamma globulin (NP16-CGG, 16 NP molecules conjugated with one CGG molecule; N-5055C-5, Biosearch Technologies) or dinitrophenyl conjugated to CpG ODN2395 (DNP-CpG, one DNP molecule conjugated with one CpG molecule, custom synthesized by Eurofins Scientific) on day 0 (100 μg in 100 μl alum, Imject Alum Adjuvant, 77161, Thermo Scientific or 50 μg in 100 μl PBS), boosted (100 μg in 100 μl PBS or 50 μg in 100 μl PBS) on day 14 and euthanized on day 28. Total, NP-specific and DNP-specific human antibodies were analyzed by specific ELISAs, as described in ‘ASCs and titration of human antibodies’. For cell sorting, NP-specific spleen huB cells from NP16-CGG-immunized THX mice were single-cell FACS sorted after staining with NP16-PE (16 NP molecules conjugated with one PE molecule, sc-396483, Santa Cruz Biotechnology; 1:100 dilution). VHDJH-CH transcripts from sorted huB cells were analyzed for SHM/CSR, B cell clonality and intraclonal diversification, as described in ‘huB cell SHM/CSR, clonality and intraclonal diversification’.

THX mice neutralizing response to Salmonella and in vivo protection

THX mice (20–24 weeks old) were injected i.p. with S. Typhimurium flagellin (CVD1925 FliC, University of Maryland School of Medicine Center for Vaccine Development, 50 μg in 100 μl alum) or nil (100 μl alum) on day 0, boosted (50 μg in 100 μl PBS or 100 μl PBS alone) on day 14 and euthanized on day 28 (ref. 39).

Total human immunoglobulin and flagellin-specific human antibodies were analyzed by specific ELISAs, as described in ‘ASCs and titration of human antibodies’. Bactericidal activity of flagellin-induced antibodies in sera from flagellin-vaccinated and non-vaccinated THX mice was measured by in vitro killing of S. Typhimurium39. S. Typhimurium IR715, a virulent nalidixic acid-resistant derivative of wild-type isolate ATCC 14028 (provided by M. Raffatellu, University of California, San Diego) was grown in LB broth (BP1426-2, Fisher Scientific) overnight at 37 °C. Log-phase cultures were prepared by diluting overnight cultures to an OD600 of 0.05 in fresh LB medium and incubating them at 37 °C, with shaking at 250 rpm until an OD600 of 0.7 or 0.8 was attained. Stock cultures were prepared by diluting 500 µl of log-phase cultures in 500 µl of 50% sterile filtered glycerol (G33-1, Fisher Scientific) then further diluted in PBS to a cell density of approximately 104 CFUs per ml. Sera from flagellin-vaccinated THX mice, non-vaccinated THX mice and healthy humans were serially twofold diluted in PBS in round-bottom 96-well plates. Diluted sera (50 µl) or PBS (50 µl, negative control) were mixed with 25 µl baby-rabbit complement (CL3441, CEDARLANE, 25% final concentration) and incubated with 25 µl diluted S. Typhimurium (250 CFUs). Each sample mixture was shaken (115 rpm) at 37 °C for 1 h and then struck onto LB-agar plates. These were incubated at 37 °C overnight, after which CFUs were enumerated. To assess the protective response induced by flagellin vaccination in vivo, flagellin-vaccinated and non-vaccinated THX mice were infected orally with S. Typhimurium (1 × 105 CFUs) by gavage on day 21. The effective dose of bacteria given to mice was verified by plating dilutions of S. Typhimurium on LB-agar plates supplemented with nalidixic acid (N8878-25G, Sigma-Aldrich, 0.05 mg ml−1). Mice were monitored for 10 days, and Kaplan–Meier survival plots were generated (GraphPad Prism v10.0.3). For cell sorting, flagellin-specific spleen huB cells from flagellin-vaccinated THX mice underwent single-cell FACS after staining with AF647-flagellin (synthesized using iLink Andy Fluor 647 Antibody Labeling Kit, L038, ABP Biosciences). VHDJH-CH transcripts from sorted huB cells were analyzed for SHM/CSR, B cell clonality and intraclonal diversification, as described in ‘huB cell SHM/CSR, clonality and intraclonal diversification’.

THX mice neutralizing antibody response to COVID-19 mRNA or RBD–KLH

THX mice (20–24 weeks old) were injected i.m. with Pfizer-BioNTech 162b2 COVID-19 vaccine (Pfizer COVID-19 mRNA, 5 µg in 50 µl PBS) or nil (50 µl PBS) on day 0, boosted (5 µg in 50 µl PBS or 50 µl PBS alone) on day 21, according to the human vaccination schedule, and euthanized on day 28. ‘Discarded’ vials of Pfizer COVID-19 mRNA vaccine were obtained from The University of Texas Health Science Center at San Antonio vaccination facility within 6 h of opening and contained less than one full vaccine dose, thereby not diverting any amount of vaccine from humans for the purpose of this study. THX mice were injected i.p. with SARS-CoV-2 Spike S1 RBD (47 amino acid peptide containing the core 37 amino acids: FRKSNLKPFERDISTEIYQAGSTPCNGVEGFNCYFPLQSYGFQPTNG, custom synthesized by ABI scientific) conjugated to KLH (RBD–KLH, 50 μg in 100 μl alum) or nil (100 μl alum) on day 0, boosted (50 μg in 100 μl PBS or 100 μl PBS alone) on day 21 and euthanized on day 28.

Total human immunoglobulin and RBD-specific human antibodies or ASCs were analyzed by specific ELISAs or ELISPOTs, as described in ‘ASCs and titration of human antibodies’. The SARS-CoV-2 neutralization power of antibodies induced by COVID-19 mRNA vaccine in THX mice was measured using two different platforms: SARS-CoV-2 Neutralizing Antibody Detection ELISA Kit (502070, Cayman Chemical) and SeroFlash SARS-CoV-2 Neutralizing Antibody Assay Fast Kit (D-1008-96, EpigenTek), according to the manufacturer’s instructions. Sera from COVID-19 mRNA-vaccinated THX mice were serially twofold diluted in PBS-Tween 20 in 96-well plates pre-coated with SARS-CoV-2 Spike S1 RBD peptide (EpigenTek platform), or a recombinant rabbit Fc-tagged SARS-CoV-2 Spike S1 RBD peptide bound to an anti-rabbit Fc-specific antibody (Cayman platform), followed by addition of recombinant His-tagged ACE2 protein to each well. These platforms use a horseradish peroxidase (HRP)-conjugated anti-His antibody in an inhibition/competition assay to measure serum neutralizing human antibody concentration by OD reading at 450 nm. High OD readings reflect a low concentration of neutralizing antibodies, while low OD readings reflect a high concentration. SARS-CoV-2-neutralizing human monoclonal antibodies were provided as a positive control by EpigenTek and Cayman. Extensive controls performed by both Cayman Chemical and EpigenTek have validated measurements of their RBD competition assays with actual virus neutralization in COVID-19-positive and COVID-19-negative human sera (https://www.caymanchem.com/product/502070/sars-cov-2-neutralizing-antibody-detection-elisa-kit; www.epigentek.com/docs/D-1008.pdf).

Sequencing and cloning of original paired heavy-chain VHDJH-CH and light-chain VκJκ-Cκ or VλJλ-Cλ gene segments for construction of human antibody-producing cell microcultures was performed by The University of Texas MD Anderson Cancer Center Recombinant Antibody Production Core. Briefly, RBD-specific spleen huB cells of three COVID-19 mRNA-vaccinated THX mice underwent single-cell FACS using biotinylated RBD peptide (47 amino acids) and FITC-streptavidin (405201, BioLegend). huVHDJH-CH and light-chain huVκJκ-Cκ or huVλJλ-Cλ gene segments from sorted huB cells were amplified as cDNAs by single-cell RT–PCR and then sequenced. The single B cell huIgH constant region and huIGκ or huIGλ constant regions were determined. The amplified huVHDJH and huVκJκ or huVλJλ cDNAs were sequenced and cloned into pcDNA3.4 vectors that included the coding sequence for either human heavy-chain (γ1) or light-chain (κ or λ) constant regions to transfect ExpiCHO cells (A29127, Thermo Fisher). Transfected ExpiCHO cells were cultured in ExpiCHO Expression Medium (A2910001, Thermo Fisher) in 100 single-cell microcultures to produce recombinant human monoclonal antibodies. After 5 days, media were collected and analyzed for RBD-specific recombinant human antibodies by specific ELISA.

ASCs and titration of human antibodies

To measure total or specific huIgM, huIgD, huIgG (huIgG1, huIgG2, huIgG3 and huIgG4), huIgA or huIgE in humanized mice, sera were diluted 400-fold or 20-fold in PBS containing 0.05% vol/vol Tween 20 (BP337-500, Fisher Scientific; PBS-Tween 20), followed by serial twofold dilution. Serially diluted samples were incubated at room temperature in 96-well plates pre-coated with goat anti-huIgM antibody (2020-01, SouthernBiotech, 1.0 µg ml−1), goat anti-huIgD antibody (2030-01, SouthernBiotech, 1.0 µg ml−1), goat anti-huIgG antibody (huIgG1, huIgG2, huIgG3 and huIgG4, 2015-01, SouthernBiotech, 1.0 µg ml−1), goat anti-huIgA antibody (2050-01, SouthernBiotech, 1.0 µg ml−1), goat anti-huIgE antibody (GE-80A, ICL Labs, 1.0 µg ml−1), NP4-BSA (four NP molecules per one BSA molecule, Biosearch Technologies, 1.0 µg ml−1), DNP5.6-BSA (average of 5.6 DNP molecules per one BSA molecule, Cosmo Bio USA, 1.0 µg ml−1, referred to as DNP5 in the Results and figure legends), BSA (Biosearch Technologies, 1.0 µg ml−1), S. Typhimurium flagellin (2.0 µg ml−1) or SARS-CoV-2 Spike S1 RBD peptide (37 amino acid core peptide, FRKSNLKPFERDISTEIYQAGSTPCNGVEGFNCYFPL, ABI Scientific, 2.0 µg ml−1) in 0.1 M sodium carbonate/bicarbonate buffer at pH 9.6. After washing plates with PBS-Tween 20, bound human antibodies were detected with biotinylated goat anti-huIgM antibody (2020-08, SouthernBiotech; 1:5,000 dilution), goat anti-huIgD antibody (2030-08, SouthernBiotech; 1:5,000 dilution), goat anti-huIgG antibody (2015-08, SouthernBiotech; 1:5,000 dilution), goat anti-huIgG1 monoclonal antibody (555869, BD Pharmingen; 1:5,000), goat anti-huIgG2 monoclonal antibody (555874, BD Pharmingen; 1:5,000 dilution), goat anti-huIgG3 monoclonal antibody (3853-6-250, MABTECH; 1:5,000 dilution), goat anti-huIgG4 monoclonal antibody (555882, BD Pharmingen; 1:5,000 dilution), goat anti-huIgA antibody (2050-08, SouthernBiotech; 1:5,000 dilution) or goat anti-huIgE antibody (9250-08, SouthernBiotech; 1:5,000 dilution; Supplementary Table 15), followed by reaction with HRP-labeled streptavidin (405210, BioLegend), development with O-phenylenediamine substrate (P8806-50TAB, Sigma-Aldrich) or 3, 3′, 5, 5′ tetramethyl benzidine substrate (421101, BioLegend), and measurement of converted substrate absorbance at 492 nm or 450 nm, respectively. Total human antibody concentrations or specific human antibody titers were calculated from OD readings (using a reference curve constructed with known antibody concentrations; BioTek Gen5 Software v2.07) and expressed as µg equivalent per ml (µg eq ml−1) or RUs (defined as the dilution factor needed to reach 50% saturation binding) using GraphPad Prism v10.0.3 software or Excel v16.83 (Microsoft) software. To measure BALF human immunoglobulin concentrations, DNP-CpG-immunized THX and huNBSGW mice were euthanized on day 28, and lungs were lavaged with 1 ml PBS containing 0.1 mM EDTA. Human immunoglobulin concentrations were measured from the recovered 1 ml lavage fluids by specific ELISA as described above.

To detect huASCs (huPBs/PCs) by ELISPOT, splenic or BM cells from DNP-CpG-immunized or COVID-19 mRNA-vaccinated THX mice were suspended in FBS-RPMI then cultured at 37 °C overnight in 96-well PVDF Multi-Screen filter plates (activated with 35% ethanol, MAIPS4510, Millipore) coated with goat anti-huIgM antibody, goat anti-huIgG antibody, goat anti-huIgA antibody, DNP5.6-BSA or SARS-CoV-2 RBD peptide (all 5 µg ml−1). Spleen and BM cells were plated at 1.25 × 105 and 2.5 × 105 cells per well to analyze total and specific huASCs, respectively. After removing supernatants, plates were incubated with biotinylated goat anti-huIgM antibody, goat anti-huIgG or goat anti-huIgA antibody for 2 h, and then, after washing, incubated with HRP-conjugated streptavidin, followed by Vectastain AEC peroxidase substrate (SK-4200, Vector Laboratories). Individual ASC spots were detected using a CTL Immunospot Analyzer and software (CTL ImmunoCapture Software v6.5.7, Cellular Technology).

Human cytokines

To measure circulating human cytokines, sera were collected from flagellin-vaccinated and COVID-19 mRNA-vaccinated THX mice and analyzed for huAPRIL, huBAFF, huIFN-γ, huIL-2, huIL-4, huIL-6, huIL-10 and huIL-21 by Luminex Human Discovery Assay 8-Plex (LXSAHM-08, R&D Systems). Analysis of huTGF-β1 was performed by TGF-β Premixed Magnetic Luminex Performance Assay (FCSTM17, R&D Systems). Samples and reagents were prepared according to the manufacturer’s instructions. Briefly, sera were diluted at a 1:2.5 (Luminex Human Discovery Assay) or 1:15 (TGF-β Luminex Performance Assay) ratio in Calibrator Diluent RD6-52 or Calibrator Diluent RD6-50, respectively. Next, 50 μl working standards and 50 μl diluted sera were each mixed with 50 μl Human Magnetic Premixed Microparticle Cocktail (color-coded magnetic beads coated with analyte-specific capture antibodies) and incubated in 96-well microplates at room temperature for 2 h with shaking at 800 rpm. After washing plates with 100 μl per well of wash buffer using a Luminex microplate magnet, human cytokines were detected by addition of 50 μl Human Premixed Biotin-Antibody cocktail (biotinylated detection monoclonal antibodies specific for analytes of interest) followed by reaction with 50 μl streptavidin–phycoerythrin and measurement using a dual-laser flow-based detection Luminex FLEXMAP 3D analyzer (Luminex). One laser classifies the beads and determines the analyte that is being detected. The second laser determines the magnitude of the PE-derived signal, which is proportional to the amount of analyte bound. Cytokine concentrations were calculated using Belysa Immunoassay Curve Fitting Software (40–122, MilliporeSigma) and compared to human physiological range93,94,95,96,97,98.

H&E, immunohistochemistry and immunofluorescence microscopyH&E and immunohistochemistry

To identify GCs in humanized mice, NP16-CGG-immunized THX, huNBSGW and JAX NSG huCD34 mouse spleens were fixed in paraformaldehyde (4%) overnight. Spleens were embedded in paraffin, sectioned, then stained with H&E or anti-huCD20 monoclonal antibody (1:200 dilution), anti-huCD3 monoclonal antibody (1:200 dilution), anti-huKi67 monoclonal antibody (1:200 dilution), anti-huBCL6 monoclonal antibody (1:200 dilution), anti-huAID (1:200 dilution) or anti-huBLIMP-1 monoclonal antibody (1:200 dilution), followed by reaction with anti-mouse IgG-HRP and brown precipitating HRP substrate 3,3′-diaminobenzidine (DAB). Spleen sectioning and staining was performed at The University of Texas Health Science Center at San Antonio Histology and Immunohistochemistry Laboratory. Images were captured using a Zeiss Imager-V.1 (ZEN Microscopy Software v3.9, 1× and 20× objective).

Immunofluorescence microscopy

To detect gut huB cells, huT cells, and huIgM-, huIgD- and huIgA-producing cells, DNP-CpG-immunized THX mouse intestines were fixed in paraformaldehyde (4%) overnight. Intestines were sectioned, then heated at 80 °C to adhere to glass slides, washed four times in xylene (214736-1L, Millipore Sigma) for 2 min, dehydrated twice with 100% ethanol for 1 min, dehydrated twice with 95% ethanol for 1 min, and washed twice in water for 1 min. Antigens were unmasked using 2 mM EDTA (15-575-020, Fisher Scientific) in 100 °C for 40 min, followed by a cooling step at 25 °C, thrice washing with TBS (15-567-027, Fisher Scientific) and final blocking by 10% BSA (BP1600-100, Fisher Scientific) for 15 min. Slides were washed again thrice with TBS and then stained with PE-Cyanine7-anti-huCD19 monoclonal antibody (clone HIB19, 302216, TONBO; 1:100 dilution), Super Bright 600-anti-huCD3 monoclonal antibody (clone OKT3, 63003741, eBioscience; 1:100 dilution), BV510-anti-huIgM monoclonal antibody (clone MHM-88, 314521, BioLegend; 1:100 dilution), BV421-anti-huIgD monoclonal antibody (clone HB-7, 348225, BioLegend; 1:100 dilution) or APC-anti-huIgA monoclonal antibody (clone IS11-8E10, 130-113-427, Miltenyi Biotec; 1:100 dilution) for 2 h in a dark, moist chamber (Supplementary Table 13). After washing thrice with 0.1% Triton X-100 (T9284, Sigma-Aldrich) in TBS, slides were air-dried, and coverslips were mounted using ProLong Gold Antifade Reagent with DAPI (P36935, Thermo Fisher Scientific). To detect human and mouse TECs, THX and huNBSGW mice thymi were snap frozen in Tissue-Tek O.C.T. Compound (45583, Sakura), sectioned by cryostat, loaded onto positively charged slides, fixed in cold acetone and stained with PE-anti-huEpCAM (ab237397, Abcam; 1:100 dilution) and PE-Cyanine7-anti-moEpCAM (118216, BioLegend; 1:100 dilution) monoclonal antibodies for 2 h at 25 °C in a moist chamber. Cover slips were then mounted on slides using ProLong Gold Antifade Reagent with DAPI. Fluorescent images were captured using a Zeiss Imager-V.1 (ZEN Microscopy Software v3.9, 20x objective).

Intestinal microbiota

Microbial DNA was extracted from feces of non-intentionally immunized THX, huNBSGW and NBSGW mice (22 weeks old) using Quick-DNA Fecal/Soil Microbe Microprep Kit (Zymo Research) according to the manufacturer’s instructions. To analyze gut bacterial microbiome composition, microbial DNA was tagged and sequenced using the Illumina MiSeq platform. Briefly, the V3–V4 hypervariable region of the bacteria 16S rRNA gene was amplified by PCR using tagged bact-341F primer 5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG-3′, bact-850R primer 5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC-3′ and Phusion high-fidelity DNA polymerase (M0530S, New England BioLabs). Multiplexing indices and Illumina sequencing adaptors were then added to the amplicons by limited-cycle amplification using the Nextera XT Index Kit (Illumina). Libraries were normalized, pooled and sequenced using the Illumina MiSeq platform. Sequencing and quality assessment were performed by The University of Texas Health Science Center at San Antonio Genome Sequencing Facility. Bacterial taxonomy was assigned using the Ribosomal Database Project (RDP) classifier v2.14 (http://rdp.cme.msu.edu/classifier/). Principle component analysis of gut bacterial composition in THX, huNBSGW and NBSGW mice was performed by ClustVis v1.0 (biit.cs.ut.ee/clustvis/), which uses clustering algorithms to construct plots visualizing similarities and/or differences between groups of samples.

Lupus THX mice, human autoantibodies, immunopathology and mortality

Lupus THX mice were generated by i.p. injection of 11 male and female THX mice (18 weeks old), constructed by huCD34+ cell engraftment of 6 NBSGW (2 males and 4 females) mice and 5 NSGW41 (2 males and 3 females) mice once with pristane (2,6,10,14-tetramethylpentadecane, P2870, Millipore Sigma, 500 μl) and continuing E2 treatment (Supplementary Table 17). Healthy THX controls (18-week-old) were constructed by huCD34+ cell engraftment of 8 NBSGW and 4 NSGW41 mice. Three additional healthy THX controls (18 weeks old) constructed by huCD34+ cell engraftment of NBSGW mice were used for ex vivo immune cell analyses and immunopathology control experiments and staining.

To measure total human immunoglobulin levels or specific human antibodies, sera from Lupus THX and control THX mice (injected with 500 µl PBS) were collected 6 weeks after pristane or PBS injection, serially twofold diluted then incubated at room temperature in 96-well plates coated with pre-adsorbed goat anti-huIgM antibody (1 µg ml−1), goat anti-huIgG antibody (1 µg ml−1), goat anti-huIgA antibody (1 µg ml−1), dsDNA (15632011, Thermo Fisher Scientific, 10 µg ml−1), histone (16736, Cayman Chemicals, 1.0 µg ml−1), Sm/RNP (A11600, Surmodics, 1.0 µg ml−1) or mouse liver RNA (10 µg ml−1). Total human antibody concentrations or specific human autoantibody titers were measured by specific ELISAs, as described in ‘ASCs and titration of human antibodies’.

To detect human antinuclear antibodies, sera from Lupus THX and healthy control THX mice, collected at 6 weeks after pristane injection, were serially diluted (from 1:50 to 1:400) in PBS and incubated on Hep-2 cell-coated slides (ANK-120, MBL-BION). Bound huIgGs were detected with FITC-anti-huIgG monoclonal antibody (clone G18-145, 555786, BD Pharmingen). Analysis of SHM/CSR, huB/huT cell clonality and intraclonal diversification in Lupus THX mice (6 weeks after pristane injection) was performed, as described in ‘huBCR IgM+ B cell and huTCR repertoires and huIgM+ B and T cell clonality’ and ‘huB cell SHM/CSR, clonality and intraclonal diversification’, in Lupus THX mice euthanized when showing obvious signs of disease and the three ‘additional’ healthy controls at corresponding ages (THX mice). To detect kidney huIgG deposition, Lupus THX and THX mice kidneys were processed for H&E and immunofluorescence staining then imaged as described in ‘H&E, immunohistochemistry and immunofluorescence microscopy’. Mortality of Lupus THX mice and THX mice was analyzed and depicted by Kaplan–Meier survival plots (GraphPad Prism v10.0.3).

Mouse IACUC and human Institutional Review Board protocols

Buffy coats were obtained upon informed consent from healthy donors, per the protocol of the South Texas Blood and Tissue Center. Human umbilical cord blood was collected from full-term, normally developed male and female newborns from healthy puerperae at the Department of Obstetrics and Gynecology, The University of Texas Long School of Medicine, The University of Texas Health Science Center at San Antonio, and obtained upon informed consent, per Institutional Review Board Protocol 17-653H. All experiments involving mice were performed in compliance with the animal protocol approved by The University of Texas Health Science Center at San Antonio Institutional Animal Care and Use Committee (IACUC protocol 20200019AR).

Sample size, randomization and statistical analysis

The exact sample size of all experiments is reported in the figure legends. In each experiment, at least five mice per group (except for the experiment of Fig. 1g) were used to ensure proper biological replicates. Sample size calculations were performed using power analysis, which accounts for effect size, standard deviation, type 1 error and 80% power in a two-sample t-test with a 5% significance level (two-sided test). G power software version 3.1.9.7 was used for these calculations. To construct humanized mice, immunodeficient mice from one litter were grafted with huCD34+ cells from the same donor. In those cases, in which litter sizes were small, multiple litters were combined and grafted with the same donor huCD34+ cells, and pups cross-fostered by a single nursing mother.

Replication: biological replicates were used in all experiments.

Randomization: After matching for sex and age, THX, huNBSGW, huNSG and JAX NSG huCD34 mice were randomly assigned to appropriate groups.

Statistical analyses: statistical analyses were performed using Excel v16.83 (Microsoft) o