Materials
CHO (Hprt−/−) cells (JCRB0218) were obtained from the Japanese Collection of Research Bioresources Cell Bank (Osaka, Japan). CHO K1 cells (RCB0285) and DT40 cells (RCB1464) were obtained from the Riken BioResource Research Center (Ibaraki, Japan). P3X63Ag8.653 myeloma cells (CRL-1580) and HCT116 cells (CCL-247) were purchased from the American Type Culture Collection (Manassas, VA, USA). TT2F cells were provided by RIKEN Centre for Biosystems Dynamics Research (Hyogo, Japan). Restriction and DNA-modifying enzymes were purchased from New England Biolabs (Ipswich, MA, USA) and TOYOBO (Tokyo, Japan), respectively. Primers were obtained from Eurofins (Huntsville, AL, USA). E. coli strains DH5α and Rosseta-gami B(DE3) pLysS were purchased from Takara Bio (Shiga, Japan) and Merck Millipore (Burlington, MA, USA), respectively.
Cell culture
A MAC vector was used to generate IGHK-NAC9. The MAC contained a mouse centromere, EGFP flanked by HS4 insulators, PGK-neo, loxP site-3′-HPRT, PGK-puro and telomeres. Chicken DT40 cells containing hChr.2 or hChr.14 were maintained at 40 °C in RPMI 1640 medium supplemented with 10% foetal bovine serum (FBS), 1% chicken serum, 50 µM 2-mercaptoethanol and 1.5 mg/mL G418. Hprt-deficient Chinese hamster ovary [CHO (Hprt−/−)] and CHO K1 cells were maintained at 37 °C in Ham’s F-12 nutrient mixture (Invitrogen, Carlsbad, CA, USA) supplemented with 10% FBS. CHO cells containing the IGHK-NAC were maintained in medium with 800 µg/mL G418. Mouse embryonic fibroblasts (MEFs) were isolated from embryos at 13.5 days postcoitum (d.p.c.). MEFs were grown in Dulbecco’s modified Eagle’s medium (DMEM) (Sigma-Aldrich, St. Louis, MO, USA) containing 10% FBS. Parental mouse ES cell line (TT2F), endogenous Ig KO ES cell subline (6TG-9) and microcell hybrid TT2F and 6TG-9 clones were maintained on mitomycin C (Sigma-Aldrich)-treated Jcl:ICR (CLEA Japan, Tokyo, Japan) MEFs and neomycin-resistant MEFs (Oriental Yeast Co., Ltd., Tokyo, Japan), respectively, as feeder layers in DMEM with 18% FBS (Hyclone Laboratories, Logan, UT, USA), 1 mM sodium pyruvate (Invitrogen), 0.1 mM non-essential amino acids (Invitrogen), 0.1 mM 2-mercaptoethanol (Sigma-Aldrich), 2 mM l-glutamine (Invitrogen) and 1000 U/mL leukaemia inhibitory factor (Funakoshi, Tokyo, Japan).
Construction of targeting vectors
Targeting vector to introduce loxP-5ʹ-HPRT into hChr.2. To prepare the homology arm, DNA from DT40 cells with hChr.2 was used as a template for PCR. A 9.5 kb homology arm was amplified using the following primers: cos138-F6B and cos138-R6B. The fragment was cloned into the BamHI site of the pKO Scrambler V901 backbone vector (Lexicon Genetics, Woodlands, TX) (V901-cos138). A PGKHyg-loxP-5ʹ-HPRT fragment obtained by AscI and KpnI digestion was cloned into the SpeI site of V901-cos138 by blunt-end ligation (pCos138HL5ʹ-H).
Targeting vector to introduce FRT-5ʹ-HPRT into hChr.2. A FRT-5ʹ-HPRT unit with cloning sites was synthesized (pkD9FRT). A CMV-Bsd fragment was inserted into pkD9FRT using EcoRI and XhoI sites to construct pBsdkD9FRT. A 4.1 kb left arm was amplified using primers kD-R9La L and kD-R9La R, and cloned into pBsdkD9FRT using NotI and MluI sites. A 3.2 kb right arm was obtained by PCR using primers kD-F9 Ra L and kD-F9Ra R, and inserted into the BamHI site (pBkD9FLR).
Targeting vector to introducing FRT-3ʹ-HPRT into hChr.14. DNA from DT40 cells with hChr.14 was used as a template for PCR to prepare the homology arm. An FRT site with cloning sites was synthesized (pSC355FRT). PGKhyg and 3ʹ-HPRT fragments were inserted into pSC355FRT using KpnI/ClaI and NheI/MluI, respectively, to construct pSC355HF3ʹH. A 3.8 kb left arm was amplified with primers NotISC355-F and AscISC355-R, and cloned into the NotI/AscI sites of pSC355HF3ʹH to construct pSC355HF3ʹHL. A 4.2 kb right arm was then prepared using primers SalISC355-F4 and BamHISC355-R4, and subcloned into pSC355HF3ʹHL to construct pSC355HF3ʹHLR. The primer sequences are described in Supplementary Table 16.
FISH
Trypsinized cells and homogenized tissue samples were incubated for 15 min in 0.075 M KCl, fixed with methanol and acetic acid (3:1), and then slides were prepared using standard methods. FISH analyses were performed using fixed metaphase or interphase spreads of each cell hybrid using digoxigenin-labelled (Roche, Basel, Switzerland) DNA [human COT-1 DNA/mouse COT-1 DNA (Invitrogen), mouse minor satellite DNA and IGK-BAC (CH17-405H5 and CH17-216K2)] and biotin-labelled DNA [human COT-1 DNA/mouse COT-1 DNA, IGK-BAC (CH17-140P2), IGH-BAC (CH17-262H11, CH17-212P11 and RP11-731F5) and each part of the targeting vector], essentially as described previously10. Chromosomal DNA was counterstained with DAPI (Sigma-Aldrich). Images were captured using an AxioImagerZ2 fluorescence microscope (Carl Zeiss GmbH, Jena, Germany).
Modification of hChr.2 and hChr.14 in DT40 cells
Homologous recombination-proficient chicken DT40 cells (1 × 107) in 0.5 mL RPMI with 25 µg of linearized targeting vector were electroporated at 550 V and 25 µF using a Gene Pulser (Bio-Rad, Hercules, CA, USA). Drug-resistant DT40 clones were selected in 1.5 mg/mL G418, 10 µg/mL blasticidin S, or 1.5 mg/mL hygromycin. Homologous recombination in DT40 hybrid clones was identified by PCR using the primers described in Supplementary Table 16.
Microcell-mediated chromosome transfer
MMCT was performed as described previously10. hChr.2-loxPFRT and hChr.14-FRT in DT40 cells were transferred to CHO (MAC) and CHO (Hprt−/−) cells, respectively, via MMCT. For each transfer, microcell hybrids were selected in medium with 800 µg/mL G418, 6 µg/mL blasticidin S, and 10 µM ouabain, and 300 µg/mL G418 and 10 µM ouabain, respectively. CHO IGK-NAC and IGHK-NAC were transferred to CHO (hChr.14-FRT) and CHO K1 cells, and selected with 600 µg/mL G418 and 4 µg/mL blasticidin S, and 800 µg/mL G418, respectively. To transfer IGHK-NAC to mouse ES cells, CHO K1 cells with IGHK-NAC were used as donor microcell hybrids. Briefly, mouse ES cells were fused with microcells prepared from donor hybrid cells and selected with G418 (250 µg/mL). The transferred IGHK-NAC in mouse ES cells was characterized by PCR and FISH.
DNA transfection
The Cre expression vector pBS185 (Invitrogen) or pCAG-FLPo was transfected into CHO hybrids with the MAC vector and modified hChr.2 or IGK-NAC and modified hChr.14 using Lipofectamine 2000 reagent (Invitrogen) in accordance with the manufacturer’s protocol. After 24 h of culture in basic growth medium, the cells were cultured in medium with 1× HAT (Sigma) and 4–6 µg/mL blasticidin S for selection. Fourteen days later, drug-resistant colonies were picked up and expanded for further analyses.
Genomic PCR
Genomic DNA was extracted from cell lines and Tc mouse tissue specimens using a genomic extraction kit (Gentra System, Minneapolis, MN, USA). PCR was then performed using the primers listed in Supplementary Table 16. Primers for hChr.2 detection were D2S177 F/R, FABP1-F/R, EIF2AK3-F/R, RPIA-F/R, IGKC-F/R, IGKV-F/R, Vk3-2 F/R and D2S159_1 F/R. Primer pairs to detect the targeted hChr.2 were cos138 sp L PAGE/cos138 sp R, x6.1 cos RA L/R, kD9 tcLa L/R and kD9 tcRa L/R. Primer pairs for hChr.14 were MTA1-F3/R3, ELK2P2-F/R, g1(g2)-F/R, CH3F3/CH4R2, and VH3-F/R. Primer pairs to detect the targeted hChr.14 were 14TarC_La F/R and 14TarC_Ra F/R. Primer pairs to detect recombination junctions were KJneo/PGKr-2, TRANS L1/R1 and PGK-r2/CMVr-1. Mouse Igκ and Igh KO were confirmed by HKD mCk L1/R1 and mCk L1/R1, and HKD mCmu L1/R1 and mCmu L2/R2, respectively. Igλ low mutation was confirmed by PCR with mIglc1 VnC L/J3C1, followed by KpnI digestion. PCR was performed using AmpliTaq Gold (PerkinElmer, Waltham, MA, USA), KOD FX (TOYOBO), or AccuPrime Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA). Amplified fragments were resolved by electrophoresis on 2% agarose gels, followed by staining with ethidium bromide.
IGHK-NAC construction
Construction of the MAC with human IGH and IGK loci employed PCR using the primers listed in Supplementary Table 16 and FISH at each step. Human chromosomes 2 and 14 were modified in homologous recombination-proficient chicken DT40 cells for recombination-mediated translocation. First, a loxP site was inserted proximally to the IGK locus on hChr.2p by homologous recombination. The targeting vector was introduced into DT40 cells with an intact hChr.2 and Neo resistance gene by electroporation and drug-resistant clones were obtained in medium with 1500 μg/mL hygromycin. FISH confirmed independent maintenance of a single copy of hChr.2 with the loxP unit (DT40 hChr.2loxP) (Supplementary Fig. 2a, b). Next, the FRT site was introduced distally to the IGK locus on hChr.2p. For loxP insertion, the targeting vector was introduced into DT40 cells with hChr.2loxP and drug-resistant clones were obtained in medium with 10 μg/mL blasticidin S. FISH confirmed independent maintenance of a single copy of hChr.2 with an FRT unit (DT40 hChr.2loxPFRT) (Supplementary Fig. 2c, d). The modified hChr.2 was transferred to CHO Hprt-/- cells with the MAC via MMCT9. Microcell hybrids were selected in medium with 800 μg/mL geneticin and 6 μg/mL blasticidin S. FISH revealed that the MAC and modified hChr.2 were independently and stably maintained in host CHO cells (Supplementary Fig. 3b).
Then, a distal region of hChr.2p from the loxP site, which included the IGK locus, was translocated to the MAC by Cre/loxP recombination (Supplementary Fig. 3a). A vector that expressed Cre under control of the CMV promoter (pBS185) was transfected by lipofection into CHO cells with the MAC and modified hChr.2. An intended reciprocal translocation between the MAC and modified hChr.2 by Cre/loxP recombination caused reconstitution of the HPRT gene in the by-product with HAT resistance, which enabled selection of CHO cell lines that carried the MAC with the IGK locus (IGK-NAC) and the by-product. Therefore, drug-resistant clones were obtained by selection in medium with 1× HAT and 4 μg/mL blasticidin S. Each recombination junction was detected by PCR and the structure of the IGK-NAC and by-product was confirmed by FISH (Supplementary Fig. 3b).
Next, we introduced a FRT site proximally to the IGH locus on hChr.14q in DT40 cells by homologous recombination (Supplementary Fig. 4a). We did not delete the distal side of the IGH locus because the IGH locus is located at the very end of hChr.14q. The targeting vector for FRT insertion was introduced by electroporation into DT40 cells that carried an intact hChr.14 with a Neo resistance gene. Drug-resistant clones were obtained by selection in medium with 1500 μg/mL hygromycin. FISH confirmed accurate targeting of hChr.14 in DT40 cells (DT40 hChr.14FRT) (Supplementary Fig. 4b). The modified hChr.14 was transferred from DT40 cells to CHO Hprt-/- cells via MMCT and microcell hybrids were obtained in selection medium with 300 μg/mL geneticin. FISH was used to confirm CHO cells with the modified hChr.14 (CHO hChr.14FRT) (Supplementary Fig. 4b). The IGK-NAC was then transferred to CHO hChr.14FRT cells by MMCT and microcell hybrids were obtained by selection in medium with 600 μg/mL geneticin and 6 μg/mL blasticidin S. FISH confirmed that the IGK-NAC and modified hChr.14 coexisted independently and stably in host CHO cells (Supplementary Fig. 5b). To clone the IGH locus into the IGK-NAC, FRT/FLP recombination-mediated reciprocal translocation between the IGK-NAC and modified hChr.14 was performed in CHO Hprt-/- cells and HPRT gene reconstruction with the desired product again enabled selection of CHO cells that carried the IGK-NAC with the IGH locus (IGHK-NAC) and the by-product. Drug-resistant clones were selected in medium with 1× HAT and 6 μg/mL blasticidin S. FISH revealed that the IGHK-NAC and by-product were independently and stably maintained in host CHO cells (Supplementary Fig. 5a, b). The resultant IGHK-NAC was transferred to CHO K1 cells to generate donor CHO K1 cells with a single desired chromosome, IGHK-NAC, for further MMCT. Microcell hybrids were selected in medium with 800 μg/mL geneticin and were monitored by GFP expression. FISH confirmed that a single copy of IGHK-NAC was independently maintained in host CHO K1 cells (Supplementary Fig. 6a, b).
TC-mAb mouse generation
To generate chimeric mice, mES cell lines were injected into eight-cell-stage embryos derived from ICR mice (CLEA, Tokyo, Japan) and then transferred into pseudopregnant ICR females. Chimeric mice with 100% coat colour chimerism were used for germline transmission. Chimeric mice were derived from an endogenous Ig KO mouse ES cell subline (6TG-9), C57BL/6 (female) × CBA (male) F1 genetic background with mouse Igh and Igκ KO, and were crossed with Jcl:ICR mice (CLEA Japan) with an ICR genetic background. F1 littermates were crossed each other or with Jcl:ICR mice. In subsequent generations, mice in the same generation were crossed with each other to produce and maintain mice with mouse Igh and Igκ KO (HKD mice). F2 mice were obtained as described above and were crossed with Crl:CD1 mice (Charles River Laboratories Japan) that have the ICR/CD-1 genetic background with Igλ low allele(s). In subsequent generations, mice in the same generation were crossed each other or with HKD mice to produce and maintain mice with mouse Igh and Igκ KO, and Igλ low alleles (HKLD mice). Chimeric mice derived from 6TG-9 mES cells with IGHK-NAC were crossed with HKD mice and offspring were further crossed with HKLD mice to generate TC-mAb mice. TC-mAb mice were maintained by crossing TC-mAb and HKLD mice. Therefore, the TC-mAb mice generated were outbred strains with a mixed genetic background derived from the ICR strain. In this study, HKLD and TC-mAb mice were of more than 10 and six generations, respectively. Resultant TC-mAb mice were used in FISH, FCM, RT-PCR and several functional assays. Representative data from these assays are shown in each figure. Jcl:ICR (RRID:IMSR_JCL:JCL:mOT-0001) and BALB/cAJcl (RRID:IMSR_JCL:JCL:mIN-0005) mice were purchased from CLEA (Tokyo, Japan) and Crl:CD1 mice were purchased from Charles River (Kanagawa, Japan). The type of animal facility was specific pathogen-free (SPF), and experimental and control animals were cohoused in a controlled ambient temperature environment with a 12-h light/dark cycle. Mice underwent isoflurane-induced anaesthesia for all blood draws and other sampling. All animal experiments were approved by the Animal Care and Use Committee of Tottori University (Permit Numbers: 14-Y-23, 15-Y-31, 16-Y-20, 17-Y-28, 19-Y-22, 20-Y-13, 20-Y-31 and 21-Y-26).
RT-PCR
Total RNA from Tc tissue specimens was prepared using ISOGEN (Nippon Gene, Tokyo, Japan), treated with RNase-free DNase I (Wako Pure Chemicals, Osaka, Japan), and purified using RNeasy columns (Qiagen, Hilden, Germany), in accordance with the manufacturer’s instructions. First-strand cDNA synthesis was performed using random hexamers and SuperScript III reverse transcriptase (Invitrogen). Primer pairs for the detection of human Ig-gene expression were as follows: Vk1BACK/Ck and CH4BACK/Cmu-111. GAPDH (RPC1/2) was used as an internal control. The primer sequences for RT-PCR analyses are described in Supplementary Table 16. cDNAs from ICR tissues were used as negative controls. PCR was performed with cDNA using AmpliTaq Gold (PerkinElmer, Waltham, MA, USA). Amplified fragments were resolved by electrophoresis on 2% agarose gels, followed by staining with ethidium bromide.
Deep sequencing analysis of Ab-coding transcripts
An NGS analysis was performed using the unbiased TCR/BCR repertoire analysis technology developed by Repertoire Genesis (Osaka, Japan). In brief, unbiased adaptor-ligation PCR was performed as previously described39. Total RNA was converted to cDNA with Superscript III reverse transcriptase (Invitrogen) and the BSL-18E primer containing polyT18 and a NotI site. Following cDNA synthesis, double-stranded (ds)-cDNA was synthesized with Escherichia coli DNA polymerase I (Invitrogen), E. coli DNA ligase (Invitrogen) and RNase H (Invitrogen). The ds-cDNA was blunted with T4 DNA polymerase (Invitrogen). A P10EA/P20EA adaptor was ligated to the 5′ end of the ds-cDNA and then cut with a NotI restriction enzyme. After elimination of the adaptor and primer with a MinElute Reaction Cleanup Kit (Qiagen), PCR was performed with KAPA HiFi DNA polymerase (Kapa Biosystems, Wilmington, MA, USA) using an IgG constant region-specific primer CG1 for BCR and P20EA. The PCR conditions were as follows: 98 °C (20 s), 65 °C (30 s) and 72 °C (1 min) for 20 cycles. The second PCR was performed with either CB2 or CG2 and P20EA primers using the same PCR conditions. Amplicons were obtained by amplification of the products from the second PCR using P22EA-ST1 and either CB-ST1-R or CG-ST1-R. The primer sequences are shown in Supplementary Table 16. Following PCR amplification, index (barcode) sequences were added by amplification with Nextera XT index kit v2 setA (Illumina Inc., San Diego, CA, USA). Equimolar concentrations of the indexed amplicon products were mixed and quantified by a Qubit 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA). Sequencing was performed using the Illumina MiSeq paired-end platform (2 × 300 bp). The human PBMC total RNA (Takara Bio USA, San Jose, CA, USA) used in this study was derived from normal human peripheral leucocytes pooled from 426 male/female Asians aged 18–54 years old.
Data analyses
All paired-end reads were classified by index sequences. Sequence assignment was conducted by determining the sequences with the highest identity in a dataset of reference sequences from the international ImMunoGeneTics information system (IMGT) database (http://www.imgt.org). Data processing, assignment and merging were executed automatically by using a repertoire analysis software program originally developed by DNA Chip Research Inc. (Tokyo, Japan).
Annotated sequence reads were defined as distinct sequence reads within the population of merged sequence reads that had been identified as a BCR gene. The copy numbers of identical annotated reads in each sample were automatically counted using the RG software program and then ranked numerically. In accordance with IMGT nomenclature, the CDR3 nucleotide sequences from a conserved cysteine at position 104 (Cys104) to a conserved phenylalanine at position 118 (Phe118) and the following glycine (Gly119) were translated to presumed amino acid sequences.
Circos analysis
Circos software was selected in this study for its high data-to-ink ratio and for its ability to clearly display relational data. Circos open-source software was acquired from www.circos.ca. The V(D)J region recombination data were reformatted using the R statistical programming language to comply with Circos data file requirements. Library sizes were normalized with Circos ideogram (circumference segments) scaling and sizing, permitting comparison of individual subgroups within libraries as well as across disparate libraries. Links, drawn from a V region to its observed J region recombinant partner, were utilized to show the frequency of recombination, with thicker links indicative of higher frequencies of recombination. The ideogram space allotted to the V region subgroup corresponds to the frequency of its observation relative to other subgroups. Analysis of V(D)J recombination was performed with an additional stacked histogram track on each Circos diagram. This track illustrates the relative proportion of each V(D)J recombination as a fraction of the total number of D region sequences observed.
Detection of somatic hypermutations in human VH and VK regions
The definition of the clone lineage referred to a set of B cells that were related by descent, which arose from the same V(D)J rearrangement event. The definition of the clonotype refers to a single Ab sequence (CDR1, 2 and 3-joined unique sequence)40. NGS reads identical to the CDR1-2-3 sequence were grouped into a single clonotype. Mutations were detected by comparison with germline sequence at every nucleotide position (around 315 nucleotides) and were calculated as a percentage. For example, when 10 reads were recorded in the same clone lineage and three reads had a point mutation at the same nucleotide position, the mutation rate at that position was described as 30%. The mutation rate at every position in the same clone lineage was integrated as a clone linage mutation rate. In addition, different lengths of Ab variable regions in annotated reads existed; therefore, an index of variable region length was set that was the average variable region length in a clone lineage and converted to 100. In this way, clone lineages could be compared at the same magnitude and the index was similar to the amino acid position in the variable region.
Diversity index
To estimate BCR diversity in deep sequence data, the Shannon-Weaver index (H′) was calculated using the following Eq. (1):
$${H}^{{prime} }=-mathop{sum }limits_{i=1}^{S}frac{{ni}}{N}{ln}frac{{ni}}{N}$$
(1)
where N is the total number of sequence reads, ni is the number of ith annotated reads and S is the species number of annotated reads41. A greater H′ value reflects greater sample diversity.
Phylogenetic analysis of the human Ab repertoire
Phylogenetic trees (circular dendrograms) were created by alignment of CDRH3 and CDRL3 amino acid sequences using the multiple sequence alignment program and the Neighbour-Joining method42. Furthermore, two phylograms of unimmunized and OVA-immunized TC-mAb mice were assembled within one phylogram based on their amino acid sequences. Additionally, copies comprising the same CDRH3 sequences were counted and overlaid on the leaves of circular dendrograms and are shown with a maximum of 50 reads; thereby, as the number of reads increased, the circle in the leaves increased.
Antigens
Ovalbumin (OVA) was obtained from Sigma (A7641). To obtain the Trx-EpEX recombinant protein, the extracellular domain of human EpCAM (NM_002354) was amplified by PCR using specific primers. The primer sequences are described in Supplementary Table 16. and subcloned into pET32b (Merck Millipore) using EcoRV and HindIII restriction enzymes (resulting in pET32b-EpEX). To obtain the GST-EpEX recombinant fusion protein, pGEX6P1 (GE Healthcare, Chicago, IL, USA) was modified by insertion of the synthesized DNA using BglII and NotI restriction enzymes (resulting in pGEX-MCS-His). The amplified EpEX fragment was also cloned into pGEX-MCS-His. To produce the AMIGO2 extracellular domain, a DNA fragment of the whole AMIGO2 region (NM_001143668) was amplified by PCR using primers and was subcloned into pET32b (Merck Millipore) using HindIII and XhoI restriction enzymes (resulting in pET32b-AMIGO2-EX). AMIGO2-EX was hard to express in Escherichia coli gami B pLysS (DE3); therefore, pET32b-AMIGO2-EX was digested with EcoRI (an EcoRI site is located near the upstream end of the leucine-rich repeats sequence), blunted using Blunting high, and then digested with EcoRV (at a site upstream of AMIGO2-EX) to eliminate the leucine-rich repeats sequence. Therefore, this vector consisted of the Ig-like domain of AMIGO2 (named pET32b-AMIGO2-Ig) and produces Trx-AMIGO2-Ig recombinant protein. After transformation of E. coli gami B pLysS (DE3) with each vector, the recombinant proteins were expressed by induction with 1.0 mM Isopropyl-β-D(-)-thiogalactopyranoside (WAKO) in LB medium. Transformation using the empty vector (pET32b) was also carried out to produce the Tag protein for use in hybridoma screening as a negative control. After harvesting cells and sonication, the recombinant proteins were obtained as inclusion-bodies. Following solubilisation with 6 M guanidine hydrochloride (WAKO) in PBS with 0.1 mM glutathione (oxide form) and 1 mM glutathione (redox form), recombinant protein was purified using Ni-NTA columns with elution using 100 mM imidazole containing 6 M guanidine hydrochloride. After dialysing the eluted fraction against PBS containing 0.4 M arginine, samples were diluted to ~1 mg/ml and stored at −30 °C. The construction of the pGEX-MCS-His vector and expression and purification of GST-AMIGO2-Ig were carried out using the same procedure as for GST-EpEX.
Immunization
Protein antigens (1 mg/ml) were prepared in PBS or in PBS containing 0.4 M arginine, and the volume corresponding to the desired amount of protein was increased to an injectable volume with PBS or PBS containing 0.4 M arginine. This volume was then mixed 1:1 (v/v) with either Freund’s or Sigma adjuvant (Sigma Adjuvant S6322; Sigma CFA F5881, Sigma) prepared in accordance with the manufacturer’s instructions. For viscous adjuvants, the solution was mixed by repeated passage through a syringe until a smooth emulsion was formed (over 30 min on ice). Injections were performed on 6-week-old male and female mice using a 1-ml glass syringe and a 27-gauge needle. Prime and boost injections were given intraperitoneally (i.p.) every 2 weeks. Volumes varied depending on the injection route and experimental requirements and were determined according to the relevant JP Home Office animal license for the procedure. Final boosts were delivered without adjuvant intravenously (i.v.) via the tail vein.
Serum concentration of Abs
The concentrations of human Igs such as hIgM, hIgG, hIgκ, hIgA and hIgE, and mouse Igs such as mIgM, mIgG, mIgκ and mIgλ were assayed using sandwich ELISA. The concentration of hIgM was assayed using a mouse monoclonal anti-human IgM Ab (Bethyl Laboratories, Montgomery, TX, USA) immobilized on 96-well plates, Nunc MaxiSorp (Thermo) and detected with peroxidase-conjugated mouse anti-human IgM Ab (Bethyl Laboratories). Similarly, hIgG, hIgκ, hIgA, hIgE, mIgM, mIgG, mIgκ and mIgλ were assayed using capture and detector Abs listed in Supplementary Table 17. The samples, standard and Ab conjugates were diluted with sample/conjugate buffer (50 mM Tris, 0.14 M NaCl, 1% BSA, 0.05% Tween 20). 3,3′,5,5′-tetramethylbenzidine (TMB) (Nacalai Tesque, Kyoto, Japan) was used as substrate, and absorbance at 450 nm was measured using a spectrophotometer (BioTek instruments, Winooski, VT, USA). The IgG subclasses were determined using an IgG Subclass Human ELISA Kit (Invitrogen) according to the manufacturer’s instructions.
Serum titre determination
Serum bleeds taken ~3 days after antigen boost were analysed by ELISA. 96-well immunoassay plates (Nunc Maxisorp) were coated with 100 µl/well of antigen at 0.5 µg/ml in PBS containing 0.4 M arginine overnight at 4 °C. Plates were washed three times with PBS-T (0.05% v/v) and blocked with PBS containing 5% skimmed milk (Difco) for 30 min at room temperature. After being washed again as above, 100 µl of serially diluted serum samples in TBS-T was added to wells and incubated for 1 h at room temperature. After incubation, plates were again washed as above and incubated with 100 µl of anti-human IgG (H + L)-HRP conjugate added at 1/50,000 dilution in TBS-T for 30 min at room temperature. Plates were washed again as above and developed using 100 µl o-phenylenediamine dihydrochloride and stopped using 25 µl 1 M H2SO4. Absorbance was read at 492 nm.
Hybridoma generation
Immunized mice were euthanized and their spleens and lymph nodes were harvested, homogenized to single-cell suspensions, and fused with myeloma P3X63Ag8.653 cells using an electro-cell-fusion generator (ECFG21) (Nepagene, Chiba, Japan). Fused hybridoma cells were seeded in 96-well plates. After ~14 days of culture, a primary screen of supernatants was performed by an ELISA. Hybridoma clones that produced EpCAM-specific Abs were identified by an ELISA using GST-EpEX following HAT selection. Cells in the positive wells were picked up and passaged in 96-well plates. Each supernatant was again analysed by the ELISA using Tag, Trx-EpEX and GST-EpEX. Hybridoma clones that reacted with Trx-EpEX and GST-EpEX, but not Tag, were established by two or more limited dilutions. In the case of Balb/c mice, hybridoma cells were screened using Tag and Trx-EpEX. Therefore, hybridoma production from TC-mAb and Balb/c mice was not performed under completely identical conditions. To produce anti-AMIGO2 mAbs, Trx-AMIGO2-Ig was used as an immunogen and GST-AMIGO2-Ig was also used to screen AMIGO2-specific mAbs.
Hybridoma screening
Hybridoma cells that produced EpCAM- or AMIGO2-specific Abs were identified by an ELISA and immunocytochemical screening. The human colorectal cell line HCT116 was used for immunocytochemical screening of anti-EpCAM mAbs. CHO cells stably transfected with human AMIGO2 were used to screen anti-AMIGO2-specific mAbs. Cultured cells were harvested from a 10-cm dish and resuspended at 2 × 105 cells/ml. Each well of a 96-well flat-bottomed plate (TPP) was seeded with 100 μl of cell suspension. Cells were incubated for 2 days at 37 °C in a CO2 incubator. The culture medium was then removed by aspiration and 100 μl of supernatant with Abs was added. After incubation for 1 h on ice, the plates were washed twice with 150 μl of the medium and 100 μl of goat anti-Human IgG (H + L) Cross-Adsorbed Secondary Ab (Abcam, Cambridge, UK) diluted at 1:400 in medium was added. Plates were washed twice with 150 μl of the medium and PBS with 1% (v/v) FBS was added. Plates were scanned under a Keyence BZ-X700 microscope.
Subclass determination
The subclasses of obtained human mAbs were determined using antigen-specific ELISA using horseradish peroxidase-conjugated secondary Abs specific for human IgG(H + L), IgG1, IgG2, IgG3, IgG4 and IgM and human Igκ and mouse Igλ. Alternatively, subclasses were determined using the Iso-Gold™ Rapid Human Antibody Isotyping Kit (BioAssay Works, Ijamsville, MD, USA) according to the manufacturer’s instructions.
Humanness score of mAbs
The amino acid sequence of obtained Abs was analysed using the T20 scoring method, which was developed to calculate the humanness of mAb variable region sequences23. A Blast search of the variable region was performed against the T20 Cutoff Human Database available at http://abanalyzer.lakepharma.com. The T20 score for an Ab is obtained from the average of the percent identities of the top 20 matched human sequences. To be considered not immunogenic, T20 scores of the FR and CDR sequences must be above 79, and T20 scores for the FR sequences only must be above 86. Scores near or above these values are predicted to be of low immunogenicity.
Surface plasmon resonance
Kinetic analysis was performed using a Biacore T200 (GE Healthcare). Each kinetic run was set up using the kinetic wizard template with six non-zero concentrations in series with at least one of the concentrations in duplicate to check the surface performance and a zero concentration. A blank immobilized surface was used as a reference surface, which was prepared as described in the ligand immobilisation step, but without any ligand. All dilutions were prepared in HBS-EP running buffer (GE Healthcare) at room temperature. Regeneration between each cycle was performed using 10 mM glycine (GE Healthcare) at pH 2.5 for 30 s. The sensor chip protein G or CM5 (GE Healthcare) was used to directly capture human Abs of interest. For kinetic analysis of anti-EpCAM Abs, five concentrations of analyte were used (10, 20, 30, 40 and 50 nM). For kinetic analysis of anti-AMIGO2 Abs, five concentrations of analyte were used (6.25, 12,5, 25, 50 and 100 nM). The data were evaluated post-run using the 1:1 kinetic binding model in Biacore T200 evaluation software to generate ka, kd, and KD43.
Cell staining and flow cytometry
To evaluate the phenotype of TC-mAb mice, we compared them with age-matched WT mice having a similar genetic background. Bone marrow and spleen tissue, and PBMCs were isolated from adult male and female mice (6–20 weeks of age) using aseptic procedures. Single-cell suspensions were prepared from the bone marrow, spleen, lymph nodes, PECs and PBMCs. Samples were stained with Abs (Supplementary Table 8) and analysed using a CytoFLEX S (Beckman Coulter, Brea, CA, USA). All staining reactions were incubated at 4 °C for 30 min using 1 × 106 cells in 100 µl staining buffer (PBS with 5% FBS:BD Biosciences Brilliant stain buffer; 1:1) (Franklin Lakes, NJ, USA) containing Mouse Seroblock FcR (Bio-Rad Laboratories, Hercules, CA, USA). Cells were stained with fluorescently labelled isotype controls (Supplementary Table 8) used to detect the positive subsets.
Immunohistochemistry
Spleens were fixed with phosphate-buffered 4% paraformaldehyde (Nacalai Tesque, Kyoto, Japan) at 4 °C for 2 h, transferred to 20% sucrose in PBS, frozen in OCT compound (Sakura Finetek) and sectioned. Frozen tissue sections on slides were permeabilized with 50 mM Tris-HCL containing 0.1% Triton-X (pH 8.0) at RT for 10 min and then blocked with Blocking One Histo (Nacalai Tesque) at RT for 10 min. Sections were incubated with a 1:100 dilution of biotin-conjugated anti-CD35 mAb (8C12, BD Biosciences, San Jose, CA, USA) and a 1:100 dilution of Alexa Fluor 647-conjugated GL7 (GL7, BioLegend, San Diego, CA. USA) in TBS-T (1 x Tris-buffered saline and 0.1% Tween 20) containing 5% Blocking One Histo at 4 °C overnight. Sections were then incubated with a 1:200 dilution of Alexa Fluor 594-conjugated streptavidin (BioLegend) and 2 µM DAPI (BioLegend) in TBS-T containing 5% Blocking One Histo at 4 °C for 45 min. Coverslips were mounted with ProLong Gold Antifade reagent (Invitrogen) and sections were analysed with a Zeiss LSM700 confocal microscope (Carl Zeiss, Oberkochen, Germany).
B cell isolation, survival assay and western blotting
Splenic B cells were purified by negative selection of CD43+ cells using anti-CD43 magnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany). The purified B cells (1 × 107 cells/ml) were stimulated with 10 µg/mL anti-mouse IgM F(ab)′2 (Jackson ImmunoResearch, West Grove, PA, USA) or 10 µg/mL anti-human IgM F(ab)′2 (Jackson ImmunoResearch) and then lysed in lysis buffer [10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100 and 0.5 mM EDTA plus protease and phosphatase inhibitor cocktails (Nacalai Tesque)]. Samples were transferred to polyvinylidene difluoride membranes by electrophoresis and analysed by immunoblotting with antibodies against p-Syk (Tyr525/526) (C87C1) and Syk (D3Z1E) (Cell Signaling Technology, Danvers, MA, USA). The purified splenic B cells were cultured with or without 25 ng/ml BAFF (R&D Systems, Minneapolis, MN, USA) for 24, 48, or 72 h. The frequency of live B cells was assessed using TOPRO3 (Invitrogen) exclusion.
Repertoire analysis of germinal centre B cells
Splenic cells were harvested on days 14 and 21 after immunization with OVA and germinal centre B cells (CD19+CD38lo/-GL7+) were sorted by a Moflo XDP (Beckman coulter). More than 1.2 million qualified reads were accumulated from each sample and assembled into merged reads. IgBlast-annotated reads were collated into datasets for subsequent analyses. Saturation of clonotype variations was confirmed in the rarefaction curve of each sample. Additionally, the annotated reads showed that >89% of the Igh and Igk transcripts in germinal centre B cells of TC-mAb mice were productive.
To estimate the degree of accumulation of somatic hypermutations, the diversity of the germline sequence of the V-segment was analysed. First, the V-D-J clone lineage was grouped by the combination of V-D-J segment typing. To compare two samples (e.g., immunized [sample 1] vs. unimmunized [sample 2]), we excluded VDJ cone lineages in which either sample had <100 clones (<100 NGS reads) because the resolution of diversity estimation was insufficient for differentiation analysis. For each V-D-J clone lineage, the statistical significance of the mean difference for %diversity of the V-segment from two samples was calculated by Welch’s t test. By setting a P-value cutoff of 0.01, the V-D-J clone lineage was categorized as follows: group 1, significantly more diverged in sample 1 M; group 2, more diverged in sample 1, but not significant; group 3, significantly less diverged in sample 1; group 4, less diverged in sample 1, but not significant. The odds score was estimated from a 2 × 2 table44. The odds score was calculated by the formula: (number of group 1/number of group 2)/(number of group 3/number of group 4). A higher odds score represented a trend of differential SHM accumulation for entire V-D-J lineages.
Statistical analysis
Statistical analyses were performed using two-tailed unpaired Student’s t test. Differences with P-values of <0.05 were considered significant. *P < 0.05.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
