We have produced mice that carry the human Ig heavy (IgH) and both κ and λ light chain transloci in a background in which the endogenous IgH and κ loci have been inactivated. The B lymphocyte population in these translocus mice is restored to about one-third of normal levels, with preferential (3:1) expression of human λ over human κ. Human IgM is found in the serum at levels between 50 and 400 μg/ml and is elevated following immunization. This primary human Ab repertoire is sufficient to yield diverse Ag-specific responses as judged by analysis of mAbs. The use of DH and J segments is similar to that seen in human B cells, with an analogous pattern of N nucleotide insertion. Maturation of the response is accompanied by somatic hypermutation, which is particularly effective in the light chain transloci. These mice therefore allow the production of Ag-specific repertoires of both IgM,κ and IgM,λ Abs and should prove useful for the production of human mAbs for clinical use.

The effective clinical application of mAb in the treatment of disease is limited by the development of an anti-Ig response in the patient. The response to foreign mAbs can be minimized by the engineering of chimeric molecules that replace the structural domains with equivalent regions from human Abs and retain only the Ag binding domains of the rodent Ab (1, 2). The disadvantage of this “humanization” process is that it must be repeated for each new Ab, placing practical limitations on the variety of target molecules that can be developed for therapy. Synthetic repertoires of human Ab fragments have been developed using combinatorial phage libraries (3); however, Ag-specific isolates from these may require further manipulation before they can provide the functional activity of mAbs from conventional hybridomas.

The alternative to these artificial libraries is to create transgenic animals carrying human Ig loci in germline configuration in the presence of defective endogenous Ig loci. The aim is that the introduced human loci function instead of the mouse loci and are rearranged and expressed to produce a fully human Ab repertoire that can be exploited to yield human mAbs of desired specificities. Early efforts using plasmid-based mini loci have revealed the feasibility of the approach, although the structural diversity of the human Ab repertoires produced is restricted owing to the limited number of germline V gene segments included in these miniloci (4, 5, 6, 7). More recent experiments using yeast artificial chromosome (YAC)5 technology has demonstrated that large regions of human DNA can be introduced into the mouse germline, allowing the production of human Ab repertoires that derive from increased number of germline V gene segments (8, 9, 10, 11, 12, 13). To date, such repertoires of wholly human Abs are only composed of human IgH,κ Abs. However, in humans (as opposed to mice) a large component of the functional Ab repertoire is provided by the Igλ locus, with Igλ-containing Ab accounting for almost half the serum Ig (14). Here we describe the production of mice that carry YAC-based human IgH, human Igκ, and human Igλ transloci in a background in which endogenous mouse IgH and Igκ chain expression has been inactivated. The introduced human transloci function to yield a fully human Ab repertoire that can be exploited to produce human mAbs of desired specificities.

The production of transgenic mice containing either the HuIgκ YAC or the HuIgλ YAC has been described previously (9, 14). The HuIgH YAC (12, 13) was modified so as to incorporate two copies of a neomycin resistance cassette in the acentric YAC arm using homologous recombination in yeast as previously described (15). The modified YAC was transferred into HM-1 ES cells (16) by protoplast fusion (15), and a clone carrying a complete copy of the IgH YAC, determined by PCR and Southern blot (data not shown), was used to derive mice following injection into BALB/c blastocysts and implantation into foster animals (8, 17). Mice with their endogenous H chain or κ L chain loci rendered nonfunctional have been described previously. The μMT−/− modification (18) inserts a stop codon and neo cassette into the membrane exon of the IgM C region, preventing the surface Ig expression during B cell development, while the MoIgκ−/− modification (19) disrupts Igκ expression by the insertion of a neo cassette in the κ C region. Transgenic animals were crossed with μMT−/− and MoIgκ−/− animals to produce mice expressing human IgM,κ Ab (referred to as four-feature mice) or both IgM,κ and IgM,λ Ab (referred to as five-feature mice). The transgenic status of the offspring was confirmed by Southern hybridization of genomic DNA with probes for the following: human IgM C region exons 1, 2, and 3; the human κ C region; and the human λ3 C region.

Flow cytometric analysis of translocus-derived Ig expression on the surface of spleen cells was performed using standard techniques. B lymphocytes were identified using B220-APC (PharMingen, San Diego, CA). Human IgM was detected using anti-human Igμ-PE (PharMingen). Human Igκ was detected using biotinylated anti-Igκ (Zymed, South San Francisco, CA) followed by streptavidin-PerCP (Becton Dickinson, Mountain View, CA), and human Igλ was detected using anti-Igλ-FITC (Sigma-Aldrich, Poole, U.K.). Mouse Igλ was detected using anti-mouse Igλ-FITC (PharMingen). Stained cells were analyzed with a FACSCalibur flow cytometer (Becton Dickinson). Data were collected on 100,000 events and were analyzed using CellQuest software (Becton Dickinson). For sorting, Peyer’s patches were isolated, and germinal center B cells were stained with B220-APC and peanut agglutinin-FITC (Sigma-Aldrich). Up to 100,000 double-positive cells were isolated for RNA preparation followed by RT-PCR of H chain and L chain rearrangements.

ELISAs to determine the serum concentrations of translocus-derived IgM, Igκ, and Igλ Ig chains were performed as described previously (4, 9, 14). Ig concentrations were standardized using purified human IgM (Sigma-Aldrich) or purified rheumatoid factor Ab prepared in the Laboratory for Molecular Recognition (Cambridge, U.K.).

Four- or five-feature mice were initially immunized with 50 μg of Ag in CFA and were boosted at 4 and 8 wk with 50 μg of Ag in IFA. A final boost was given at 12 wk, and 3 days later hybridomas were prepared by fusion of splenocytes with NS/0 myeloma cells using polyethylene glycol 1500. Fusion supernatants were screened for reactivity with the immunogen by ELISA, and selected colonies were expanded for further analysis. Human IgM expression levels and L chain use were determined by ELISA. The specificity of the hybridomas was confirmed by testing for cross-reactivity to unrelated Ags.

Total cytoplasmic RNA was isolated from hybridoma cells or sorted Peyer’s patch cells using Tripure reagent (Boehringer Mannheim U.K., Lewes, U.K.), following the manufacturer’s instructions, and reverse transcribed using oligo-(dT)22 and Superscript II Reverse Transcriptase (Life Technologies, Paisley, U.K.). The primers used for PCR amplification of rearranged Ig genes are listed in Table I. Rearranged human Ig H chains were amplified using the family-specific leader primers and the IgM Constant primer (20), with separate reactions set up for each leader primer. Rearranged Igκ genes were amplified using the Vκ primer and the κ C primer (21). Rearranged λ genes were amplified using the Vλ2 and Vλ3 in separate reactions with the λC primer (14). The amplified rearrangements were purified using the QIAquick Prep System and cloned into the pGem-T vector (Promega U.K., Southampton, U.K.). Recombinant colonies were screened for appropriate rearrangements by PCR using primers to the respective framework 1 and framework 4 regions (Table I). Plasmid was isolated and sequenced using standard M13 forward or reverse sequencing primers.

Table I.

Primers for amplification of human Ig gene rearrangements from translocus mice

Sequence
Heavy chain family leaders (sense, codons−20 to−13)  
VH1 leader 5′-ATG GAC TGG ACC TGG AG-3′ 
VH2 leader.1 5′-ATG GAC ATA CTT TGT TCC ACG C-3′ 
VH2 leader.2 5′-ATG GAC ACA CTT TGC TCC ACG C-3′ 
VH4 leader 5′-ATG AAA CAC CTG TGG TTC TTC-3′ 
VH6 leader 5′-ATG TCT GTC TCC TTC CTC ATC-3′ 
Heavy chain constant (antisense, codons 127–134)  
IgM constant 5′-CGT ATC CGA CGG GGA ATT CTC ACA-3′ 
Heavy chain consensus framework 1 (sense, codons 1–8)  
Universal VH 5′-GAG GTG (AC)A(AG) CTG CAG (CG)AG TC(AT) GG-3′ 
Heavy chain consensus framework 4 (antisense, codons 103–111)  
Heavy joining 5′-CAG GGT GAC CAG GGT ACC TTG GCC CCA G-3′ 
κ light chain consensus framework 1 (sense, codons 1–8)  
Vκ all 5′-GA(AC) A(CT)(CT) GAG CTC ACC CAG TCT CCA-3′ 
κ light chain constant (antisense, codons 112–119)  
κ constant 5′-CGG GAA GAT GAA GAC AGA TGG TGC-3′ 
κ light chain consensus framework 4 (antisense, codons 101–108)  
κ join 5′-G TTT GAT CTC CAG CTT GGT CCC-3′ 
λ light chain family framework 1 (sense, codons 1–7)  
Vλ2 5′-CAG TCT GCC CTG ACT CAG CCT-3′ 
Vλ3 5′-TCC TAT GAG CTG AC(AT) CAG-3′ 
λ light chain constant (antisense, codons 126–132)  
λ constant 5′-CG TGT GGC CTT GTT GGC T-3′ 
λ light chain consensus framework 4 (antisense, codons 101–106a)  
λ join 5′-TAG GAC GGT (CG)A(CG) CTT GGT CCC-3′ 
Sequence
Heavy chain family leaders (sense, codons−20 to−13)  
VH1 leader 5′-ATG GAC TGG ACC TGG AG-3′ 
VH2 leader.1 5′-ATG GAC ATA CTT TGT TCC ACG C-3′ 
VH2 leader.2 5′-ATG GAC ACA CTT TGC TCC ACG C-3′ 
VH4 leader 5′-ATG AAA CAC CTG TGG TTC TTC-3′ 
VH6 leader 5′-ATG TCT GTC TCC TTC CTC ATC-3′ 
Heavy chain constant (antisense, codons 127–134)  
IgM constant 5′-CGT ATC CGA CGG GGA ATT CTC ACA-3′ 
Heavy chain consensus framework 1 (sense, codons 1–8)  
Universal VH 5′-GAG GTG (AC)A(AG) CTG CAG (CG)AG TC(AT) GG-3′ 
Heavy chain consensus framework 4 (antisense, codons 103–111)  
Heavy joining 5′-CAG GGT GAC CAG GGT ACC TTG GCC CCA G-3′ 
κ light chain consensus framework 1 (sense, codons 1–8)  
Vκ all 5′-GA(AC) A(CT)(CT) GAG CTC ACC CAG TCT CCA-3′ 
κ light chain constant (antisense, codons 112–119)  
κ constant 5′-CGG GAA GAT GAA GAC AGA TGG TGC-3′ 
κ light chain consensus framework 4 (antisense, codons 101–108)  
κ join 5′-G TTT GAT CTC CAG CTT GGT CCC-3′ 
λ light chain family framework 1 (sense, codons 1–7)  
Vλ2 5′-CAG TCT GCC CTG ACT CAG CCT-3′ 
Vλ3 5′-TCC TAT GAG CTG AC(AT) CAG-3′ 
λ light chain constant (antisense, codons 126–132)  
λ constant 5′-CG TGT GGC CTT GTT GGC T-3′ 
λ light chain consensus framework 4 (antisense, codons 101–106a)  
λ join 5′-TAG GAC GGT (CG)A(CG) CTT GGT CCC-3′ 

Sequence data were aligned to the germline V, D, and J region sequences known to be on the YAC constructs using MacVector software (Oxford Molecular, Oxford, U.K.), and the segments used by the rearrangement and any point mutations present were identified. The nomenclature for the various gene segments follows a family- and position-based scheme, as described in the IMGT database (22).

The three different transloci are illustrated in Fig. 1. For the human heavy chain translocus, we used a 240-kb YAC (HuIgH) that contains the core region of the human IgH locus comprising five VH segments and the complete DH and JH loci linked to Cμ-δ in correct germline configuration; the isolation and characterization of this YAC have been previously described (13). The HuIgH YAC was modified by insertion of a neomycin resistance gene into the acentric YAC arm and introduced into embryonic stem cells using the protoplast fusion technique as described in Materials andMethods. Mice carrying the HuIgκ and HuIgλ transloci have been previously described (9, 14). The 1.3-Mb HuIgκ YAC (9) contains a complement of 103 Vκ region gene segments comprising 20 repeats of five Vκ region genes obtained from the cosmid cos106 (23), attached to the core of the germline locus, including three Vκ regions, the complete Jκ cluster, Cκ, and the κ deleting element. One of the repeated Vκ region segments (Vκ2D-10) and one from the core region (Vκ7-3) are pseudogenes, making a complement of 82 functional Vκ segments in the locus. The 410-kb HuIgλ YAC (24) contains 28 Vλ region gene segments (with 16 functional segments) and the seven paired Jλ and Cλ segments (four of which are functional), in the correct germline configuration.

FIGURE 1.

Diagrammatic maps of the HuIgHeavy, HuIgκ, and HuIgλ YACs used to produce human Ig transgenic mice. The positions of the functional elements are indicated by boxes. V regions are numbered according to their gene family and their position in the locus, following the system in Ref. 22 . The HuIgHeavy YAC contains the complete D and J region loci, the intron enhancer (not marked), and the Igμ and Igδ C regions. The HuIgκ YAC includes the five κ J regions, the κ C region, the 3′ enhancer, and the κ deleting element. The HuIgλ YAC contains the seven paired λ J and C regions, four of which are functional, and the 3′ enhancer. The scale is indicated by the 100-kb bar.

FIGURE 1.

Diagrammatic maps of the HuIgHeavy, HuIgκ, and HuIgλ YACs used to produce human Ig transgenic mice. The positions of the functional elements are indicated by boxes. V regions are numbered according to their gene family and their position in the locus, following the system in Ref. 22 . The HuIgHeavy YAC contains the complete D and J region loci, the intron enhancer (not marked), and the Igμ and Igδ C regions. The HuIgκ YAC includes the five κ J regions, the κ C region, the 3′ enhancer, and the κ deleting element. The HuIgλ YAC contains the seven paired λ J and C regions, four of which are functional, and the 3′ enhancer. The scale is indicated by the 100-kb bar.

Close modal

Four-feature human Ig-expressing mouse strains were derived by crossing of strains carrying the HuIgH YAC and the μMT−/− knockout with mice carrying the HuIgκ YAC and the Moκ−/− knockout (9). The remaining HuIgλ translocus was introduced by crossing the four-feature mice with mice carrying the HuIgλ YAC and the Moκ−/− knockout (14), with care to retain the μMT−/− knockout. The strains were bred to carry two alleles of each of the transloci. Test breeding showed that the three transloci and two knockouts were not linked, indicating that the integration of the YACs was random and independent.

Flow cytometric analysis of spleen populations was conducted to test whether the human transloci were capable of rescuing B cell development in the μMT−/−, Moκ−/− background. The percentage of B lymphocytes (B220-positive) in the spleens of nonimmunized four-feature and five-feature mice maintained in barrier conditions ranged from 5–25%; representative animals are shown in Fig. 2. This is compared with spleens from normal BALB/c mice, which contain ∼40–45% B220-positive cells, and μMT−/− mice, which contain <2% B220-positive cells (Fig. 2 A). This level of B cell reconstitution is equal to or greater than what has been reported for other human Ig mice (6, 11).

FIGURE 2.

Flow cytometric analysis of B lymphocytes from spleens of four- and five-feature mice. A, Percentage of B cells in spleens of wild-type, μMT−/− and representative four- and five-feature mice. Essentially no B cells are present in the μMT−/− mice, and the introduction of the human Ig YACs restores the B220+ cell population in four- and five-feature mice to 25–45% of wild-type levels. B, Percentages of B220+ cells (gated as in A and comparing the numbers of B220+ cells in 100,000 spleen cells) expressing surface human Igκ and mouse Igλ (four feature, upper panels) or surface human Igκ and human Igλ (five feature, lower panels). In the four-feature mice the κ translocus is dominant over the endogenous λ locus, while in the five-feature animals the λ translocus is dominant over the κ translocus.

FIGURE 2.

Flow cytometric analysis of B lymphocytes from spleens of four- and five-feature mice. A, Percentage of B cells in spleens of wild-type, μMT−/− and representative four- and five-feature mice. Essentially no B cells are present in the μMT−/− mice, and the introduction of the human Ig YACs restores the B220+ cell population in four- and five-feature mice to 25–45% of wild-type levels. B, Percentages of B220+ cells (gated as in A and comparing the numbers of B220+ cells in 100,000 spleen cells) expressing surface human Igκ and mouse Igλ (four feature, upper panels) or surface human Igκ and human Igλ (five feature, lower panels). In the four-feature mice the κ translocus is dominant over the endogenous λ locus, while in the five-feature animals the λ translocus is dominant over the κ translocus.

Close modal

Human Igκ L chain was expressed on the surface of 45–65% of B cells from spleens of four-feature animals, with 16–25% of B cells expressing mouse Igλ, a κ:λ ratio of ∼3:1 (Fig. 2,B). In five-feature animals, 10–20% of B cells expressed human κ L chain, while 45–60% expressed human λ, a κ:λ ratio of 1:3 (Fig. 2 B). The level of mouse Igλ+ B cells in the five-feature animals was <5%. A small number of L chain double-positive spleen cells were detected, which probably means that (as described in normal mice (Ref. 14 and references therein)) there is a low degree of leakage in isotype exclusion at the light chain loci. Human peripheral B lymphocytes typically have a κ:λ ratio of 3:2, while in normal mice the ratio is typically 19:1. The high contribution of the HuIgλ YAC to the Ig repertoire was also seen in mice containing the HuIgλ YAC in the presence of a functional mouse Igκ locus (up to 40% of the splenic B cells expressing human λ) (14) and is in contrast to the low expression of the HuIgκ locus to the Ig repertoire when a functional mouse Igκ locus is present (with up to 15% of B cells expressing human κ L chain on the cell surface) (9).

In the four- and five-feature mice, human IgM was present at between 50 and 400 μg/ml (Fig. 3). The level varied between individual animals and tended to be higher in the five-feature animals, but was not correlated to the number of B cells present in the spleen in these animals. The increased serum IgM levels in five-feature mice may result from higher secretion by cells expressing human IgM,λ Ab due to the stronger transcriptional activity of the Igλ 3′ enhancer (25). Normal BALB/c mice maintained under pathogen-free conditions have serum IgM levels of ∼500 μg/ml and total IgG levels of ∼400 μg/ml (26). The level of human IgM in the translocus mice is similar to what was found in mice made using plasmid-based miniloci (6, 27), although these smaller loci contain fewer germline segments to form a diverse repertoire. The distribution of human L chain in the serum of five-feature mice paralleled what was found on the surface of spleen B cells, with Igκ levels 3- to 4-fold lower than human Igλ levels (data not shown).

FIGURE 3.

Human IgM expression in serum of four- and five-feature mice. A, Levels in serum of representative four-feature mice (light panels) and five-feature mice (dark panels) kept under barrier conditions, in which the IgM concentration was usually greater in serum from five-feature animals. The four-feature mice 6130 and 6133 and five-feature mice 5079 and 5086 are the animals analyzed for Fig. 2 above. The human IgM background level in normal mice is <0.1 μg/ml (27 ). B, Elevation of human IgM in serum of a four-feature mouse following immunization, performed as described in Materials and Methods. C, Specific anti-progesterone-BSA response following immunization of a four-feature mouse. The Ag-specific response was determined by coating the ELISA plate with 3 μg/ml progesterone-11-BSA in PBS. Ag-specific hybridomas 1451-A5 and 1451-B9 (Table II and Fig. 4) were cloned following fusion of the spleen of this mouse.

FIGURE 3.

Human IgM expression in serum of four- and five-feature mice. A, Levels in serum of representative four-feature mice (light panels) and five-feature mice (dark panels) kept under barrier conditions, in which the IgM concentration was usually greater in serum from five-feature animals. The four-feature mice 6130 and 6133 and five-feature mice 5079 and 5086 are the animals analyzed for Fig. 2 above. The human IgM background level in normal mice is <0.1 μg/ml (27 ). B, Elevation of human IgM in serum of a four-feature mouse following immunization, performed as described in Materials and Methods. C, Specific anti-progesterone-BSA response following immunization of a four-feature mouse. The Ag-specific response was determined by coating the ELISA plate with 3 μg/ml progesterone-11-BSA in PBS. Ag-specific hybridomas 1451-A5 and 1451-B9 (Table II and Fig. 4) were cloned following fusion of the spleen of this mouse.

Close modal

Following immunization and boosts, the total serum IgM concentration was elevated, and specific Ab could be detected (Fig. 3). The titer of specific IgM produced was lower than that seen during parallel immunization of BALB/c mice (data not shown). Hybridomas secreting Ag-specific human IgM,κ Ab were obtained from four-feature mice following immunization with the Ags listed in Table II. In culture, the hybridomas secreted ∼2 μg of human IgM/ml of supernatant, compared with levels of 10 μg/ml or more from mouse IgM hybridomas grown in parallel.

Table II.

Ig rearrangements of specific hybridomas from four-feature human Ig mice

ImmunogenClone NameVH-DH-JH SegmentsVκ-Jκ Segments
PLAP 1414 /20 VH1.2-DH2.2-JHVκ1D.12-Jκ4 
Human IgE 1431 /AB5 VH1.2-DH3.16-JHVκ1D.12-Jκ1 
Prog-BSA 1451 /A5 VH1.2-DH1.7-JHVκ4.1-Jκ2b 
Prog-BSA 1451 /B9 VH1.2-DH1.7-JHVκ1D.12-Jκ4 
Prog-BSA 1477 /13 VH1.2-DH4.4/11-JHVκ4.1-Jκ5 
Prog-BSA 1477 /14 VH1.2-DH?-JHVκ4.1-Jκ4 
IGF-1 1559 /3 VH1.2-DH3.16-JHVκ3D.11-Jκ1 
ImmunogenClone NameVH-DH-JH SegmentsVκ-Jκ Segments
PLAP 1414 /20 VH1.2-DH2.2-JHVκ1D.12-Jκ4 
Human IgE 1431 /AB5 VH1.2-DH3.16-JHVκ1D.12-Jκ1 
Prog-BSA 1451 /A5 VH1.2-DH1.7-JHVκ4.1-Jκ2b 
Prog-BSA 1451 /B9 VH1.2-DH1.7-JHVκ1D.12-Jκ4 
Prog-BSA 1477 /13 VH1.2-DH4.4/11-JHVκ4.1-Jκ5 
Prog-BSA 1477 /14 VH1.2-DH?-JHVκ4.1-Jκ4 
IGF-1 1559 /3 VH1.2-DH3.16-JHVκ3D.11-Jκ1 

The distribution of L chain expression by clones obtained following immunization also paralleled the distribution on the surface of spleen cells. In one fusion from a four-feature animal, human Igκ was expressed by 14 of 16 hybridoma clones, with the remaining clones expressing mouse Igλ. In seven fusions from five-feature animals, more human Igλ-expressing clones were obtained on four occasions, with from 4- to 8-fold more human Igλ-expressing clones isolated. In the remaining fusions, the ratio of human Igκ to human Igλ-expressing clones was almost 1:1. Combined L chain expression from different loci, i.e., a combination of human κ with human λ or mouse λ L chains, was not observed.

The H chain and L chain rearrangements from the specific hybridomas described above were sequenced to determine the diversity of the immune response. The H chains showed a very restricted use of the V region segments, dominated by VH1-2, accompanied by a highly diverse CDR3 rearrangement (Table II and Fig. 4). The use of the L chain V segments was not so restricted, and this was coupled with the use of different J regions. The potential diversity of the response to a single Ag was shown by the several hybridomas raised against progesterone-BSA (1451/A5, 1451/B9, 1477/13, and 1477/14), all of which had unique H chain and L chain rearrangements. To further assess the diversity of the available Ab repertoire generated in the translocus mice, a total of 41 H chain rearrangements, 19 κ L chain rearrangements, and 6 λ L chain rearrangements were isolated by RT-PCR from hybridomas, spleen cells, and Peyer’s patch germinal center cells. All the sequences isolated used translocus-derived segments in a rearrangement that was in the correct reading frame to produce a functional Ab protein.

FIGURE 4.

Nucleotide sequences of the CDR3 of translocus-derived H chain and κ L chain rearrangements from the Ag-specific hybridomas listed in Table II, indicating the V, D, and J segments used. All rearrangements were productive and are presented showing the correct reading frame. The codon numbers are as defined in Ref. 47 , with the H chain CDR3 being from codons 95–102 inclusive, and the κ L chain CDR3 from codons 89–97 inclusive. A, Junctions of H chain rearrangements showing from codon 92 (framework 3) to codon 105 (framework 4). D and J segments are indicated by underlining. Potential N nucleotides are indicated using lowercase letters. If the D and J segments overlapped, priority in assignment was given to the J segment. B, Junctions of κ L chain rearrangements shown from codon 86 (framework 3) to codon 102 (framework 4). The portion of the CDR3 encoded by the V region and Jκ is indicated by underlining. Potential N nucleotides are indicated by lowercase letters.

FIGURE 4.

Nucleotide sequences of the CDR3 of translocus-derived H chain and κ L chain rearrangements from the Ag-specific hybridomas listed in Table II, indicating the V, D, and J segments used. All rearrangements were productive and are presented showing the correct reading frame. The codon numbers are as defined in Ref. 47 , with the H chain CDR3 being from codons 95–102 inclusive, and the κ L chain CDR3 from codons 89–97 inclusive. A, Junctions of H chain rearrangements showing from codon 92 (framework 3) to codon 105 (framework 4). D and J segments are indicated by underlining. Potential N nucleotides are indicated using lowercase letters. If the D and J segments overlapped, priority in assignment was given to the J segment. B, Junctions of κ L chain rearrangements shown from codon 86 (framework 3) to codon 102 (framework 4). The portion of the CDR3 encoded by the V region and Jκ is indicated by underlining. Potential N nucleotides are indicated by lowercase letters.

Close modal

Further analysis confirmed this restriction in H chain variability. Of 41 H chain rearrangements, 36 used the VH1-2 V gene, with VH6-1 and VH4-4 being found rarely. The other genes, VH1-3 and VH2-5, were not found in the sequences analyzed, and rearrangements using the VH2-5 segment could not be detected by RT-PCR from splenic B cells (data not shown). For the H chain rearrangements, the source of repertoire variability is the CDR3 region formed by the selection of D and J segments and by the V-D and D-J junctions. The length of the CDR3 region ranged from 7–21 aa (Fig. 5), with a distribution similar to H chain rearrangements from human B lymphocytes (28). The use of D segments is of particular interest, as the mouse D locus is more restricted than the human locus. The mouse locus has only 12 D segments, divided into two main families, DSP2 and DFL16, together with the DQ52 segment (29), while the human D locus contains 27 segments, divided into six families along a 40-kb region (30). D segments could be identified in 39 of the 41 sequences, with 13 of the 27 segments present in the locus found in rearrangements (Fig. 6). The majority of the D segments used were from the 5′ region of the locus (further from the J segments), which parallels D segment utilization in human B lymphocytes (28). The use of H chain J segments is also similar to what is seen in human rearrangements, with the JH4 (18 of 41) and JH6 (19 of 41) segments being used most often. Nongermline-encoded N nucleotides were found in all rearrangements, with 26 of the 39 sequences containing D segments having insertions at both the V-D (average, 5.2 ± 2.9 bases) and D-J (average, 2.3 ± 2.7 bases) junctions. The number of N nucleotides added in the H chain is intermediate between the numbers reported for human V-D (average, 7.3 bases) and D-J (6.3 bases) joins and those for mouse V-D (average, 3.2 ± 2.5 insertions) and D-J (average, 2.2 ± 2.3) joins (30, 31). A similar level of N nucleotide insertion in rearranged H chain sequences was observed in other strains of transgenic mice expressing human Ig repertoires (10, 11).

FIGURE 5.

Distribution of the CDR3 length of H chain rearrangements from four- and five-feature mice (a total of 41 sequences) and from human lymphocytes (data from Ref. 28 ). The H chain CDR3 is defined as from codons 95–102 inclusive (see Fig. 4).

FIGURE 5.

Distribution of the CDR3 length of H chain rearrangements from four- and five-feature mice (a total of 41 sequences) and from human lymphocytes (data from Ref. 28 ). The H chain CDR3 is defined as from codons 95–102 inclusive (see Fig. 4).

Close modal
FIGURE 6.

Occurrence of individual D segments in H chain rearrangements from four- and five-feature mice (a total of 39 sequences) and from human lymphocytes (data from Ref. 30 ). D segments are numbered according to their family and position in the locus (30 ), with segment D1-1 being furthest from, and D7-27 being nearest to the J regions. D4-4 and D4-11 have identical coding sequences, and the appearance of this sequence in rearrangements has been divided equally between the two segments.

FIGURE 6.

Occurrence of individual D segments in H chain rearrangements from four- and five-feature mice (a total of 39 sequences) and from human lymphocytes (data from Ref. 30 ). D segments are numbered according to their family and position in the locus (30 ), with segment D1-1 being furthest from, and D7-27 being nearest to the J regions. D4-4 and D4-11 have identical coding sequences, and the appearance of this sequence in rearrangements has been divided equally between the two segments.

Close modal

The Igκ rearrangements used the Vκ4–1 and Vκ1D-12 segments most frequently (7 of 19 and 9 of 19, respectively), with Vκ3D-11 and Vκ1D-13 being used in only a few rearrangements. The other functional Vκ genes in the construct, Vκ5–2 and Vκ1D-9, were not found in the sequences examined. The six Igλ rearrangements were obtained from a single animal and show a limited use of the V gene segments, with Vλ3–19 used in five sequences. Given the high contribution to the B cell repertoire seen in FACS and serum analysis, it is likely that the rearrangement of the locus in the five-feature mice is similar to what is seen in mice in which HuIgλ YAC is in the presence of a functional mouse Igκ locus (14). Junctional diversity is less obvious for the κ L chain rearrangements and was provided mainly through the use of different J segments. The CDR3 length was restricted, with 18 of the 19 Vκ-Jκ sequences encoding a 9-aa CDR3, and the remaining sequence encoding an 8-aa CDR3. The same restriction has been reported for mouse Vκ-Jκ sequences, where 39 of 41 productive Vκ-Jκ rearrangements encoded a 9-aa CDR3, with the remaining CDR3 being 10 and 11 aa (32). In contrast, human Vκ-Jκ CDR3 can range from 7–12 aa, with at least 20% of sequences encoding CDR3 longer than 9 aa (33). The increased variation in human L chain sequences is due to the insertion of N nucleotides at the V-J junction, which is not seen in L chain rearrangements from mice (32, 33). Little or no N insertion is found in the translocus-derived L chains either in four- or five-feature mice or in mice with the HuIgκ or HuIgλ YAC in the presence of a functional mouse H chain locus (14, 21). This would suggest that the L chain translocus rearranges at the same developmental stage as the endogenous L chains, at which time terminal deoxynucleotide transferase activity is reduced.

The Ig rearrangements were examined for evidence of somatic hypermutation, which would indicate whether the B lymphocytes expressing human IgM were able to participate normally in immune responses. The majority of L chain rearrangements (13 of 19 Vκ-Jκ and five of six Vλ-Jλ) contained two or more differences from the germline sequence, with up to 15 point mutations/sequence, while in contrast only four of 41 H chain rearrangements contained two or more mutations (Fig. 7). The frequency of mutations in the κ L chains was similar to what was observed in mice carrying the HuIgκ YAC in the presence of a functional mouse H chain locus (21). In the paired sequences from 19 hybridomas, six clones had no mutations in either the H chain or L chain, 10 clones had an unmutated H chain with a mutated L chain, and three clones had mutations in both chains. The most mutated H chain sequence came from the same hybridoma as the most mutated κ L chain sequence (hybridoma 1431/AB5).

FIGURE 7.

Somatic hypermutation of translocus-derived rearrangements from four- and five-feature animals. The proportion of H chain and κ and λ L chain sequences with zero, one, two, etc., mutations are shown using pie charts. The number of sequences analyzed for each locus is indicated by the number in the center of the chart. The CDR3 region was excluded from the mutation analysis, as this may be affected by the junctional diversity during rearrangement.

FIGURE 7.

Somatic hypermutation of translocus-derived rearrangements from four- and five-feature animals. The proportion of H chain and κ and λ L chain sequences with zero, one, two, etc., mutations are shown using pie charts. The number of sequences analyzed for each locus is indicated by the number in the center of the chart. The CDR3 region was excluded from the mutation analysis, as this may be affected by the junctional diversity during rearrangement.

Close modal

The distribution of the mutations in the κ L chain rearrangements shows that of the 28 changes in the CDRs, 25 lead to amino acid replacement, with a replacement:silent ratio near 8:1, while of 50 point mutations in the framework regions, 27 were replacement mutations in the encoded Ab, with a replacement:silent ratio near 1:1. The bias in favor of replacement mutations in the CDR regions is associated with selection-improved Ag binding (34). No such bias was observed for the mutations in the H chain rearrangements. These results suggest that the B cells expressing human surface Ig are capable of participating in an immune response and undergoing affinity maturation, but that the HuIgHeavy YAC is a poor target for the introduction of mutations.

The four- and five-feature human Ig mice we have produced are capable of forming large Ab repertoires and can produce a diverse response following immunization. As such, they may be useful for the production of human mAbs for clinical use. In addition, the characteristics of the B cell repertoires that are developed suggest that these mice may be useful in studying the role of Ig loci in B cell development.

The human Ig transloci are introduced into a background in which the endogenous mouse H chain and κ L chain loci are nonfunctional, with only the mouse λ locus unaltered. While it is clear that there is significant reconstitution of the B cell repertoire in these mice, the number of B cells present in the mice is reduced compared with that in mice with functional endogenous loci (Fig. 2 A). The comparison of several strains carrying different transgene and knockout combinations suggests that the low B cell pool results from poor function of the HuIgHeavy YAC in replacing the mouse H chain locus.

The HuIgκ and HuIgλ YACs are each capable of substituting for the mouse Igκ locus during B cell development, leading to the complete restoration of the B cell pool. The knockout of the κ locus (19), with a functional mouse H chain locus present, reduces the B cell population from around 40% of spleen cells to about 15%, with these cells expressing mouse λ L chains. The introduction of either the HuIgκ YAC or the HuIgλ YAC to this background restores the B cell population to near normal levels (9, 14). In contrast, the introduction of the HuIgH YAC only increases the B cell population to 5–25% of that of spleen cells, from the <2% present with the μMT−/− modification (Fig. 2 A). The failure of the introduced locus to restore the B cell repertoire is not due to competition with endogenous H chain loci, which are still able to rearrange in the μMT−/− background before the functional defect is manifest. The restoration of the B cell repertoire is also not complete if a germline configuration H chain translocus is introduced into a knockout background in which the entire J segment locus is removed, and no rearrangement of the endogenous alleles is possible (11, 35). Similarly, the introduction of a larger germline configuration H chain YAC containing the majority of functional V regions does not restore the B cell population to a greater extent than a locus containing five V regions (10, 11).

Analysis of various L chain transgenes indicates that elements in the locus downstream of the C regions may be necessary during the development of the B cell repertoire (21). No H chain translocus has been described that includes these regions in germline configuration, and those transloci that do include downstream elements, such as other C region genes or downstream enhancers (6, 10), place these elements in close proximity to the IgM and IgD domains and do not allow for the presence of any essential regulatory elements in the regions between the C domains.

The distribution of surface L chain in the five-feature animals further indicates that the structure of the translocus can affect the contribution to the B cell repertoire. The reconstitution of the B cell population was similar in both four-feature and five-feature mice (Fig. 2,A), yet the total serum IgM levels tended to be higher in the five-feature mice (Fig. 3). Although the individual transloci can both substitute for the mouse Igκ locus, the contribution of the HuIgλ locus appears to be dominant when in direct competition with the HuIgκ locus, with a κ:λ ratio of 1:3. In human B cells, Igκ is expressed slightly more frequently, with a typical κ:λ ratio of 60:40. In the human genome, both κ and λ V regions are gathered in clusters along their respective loci (36, 37), and the majority of L chain rearrangements (κ or λ) use V segments from the proximal cluster, nearer the core of the locus (38, 39). The dominant HuIgλ YAC is arranged in the correct germline configuration (24), whereas the HuIgκ YAC was produced by attaching a multimer of five V regions to the core of the human κ locus of three V regions, the J and C segments, and the enhancers (9). Both transloci contain the 3′ enhancers in their correct location, but the format of the Igκ YAC may remove other regulatory elements from upstream of the core of the locus. The absence of these elements may affect the rearrangement of the κ translocus and therefore lower its contribution to the Ig repertoire in the five-feature animals. As the mouse λ locus rearranges so poorly, even when the mouse κ locus is nonfunctional (9), this deficiency in HuIgκ YAC is not evident in the absence of HuIgλ YAC, either in the four-feature animals or when HuIgκ YAC is present with the functional mouse H chain locus.

The characteristics of the human Ig rearrangements indicate that the formation of the Ab repertoire from the translocus elements follows the same rules as that for the equivalent mouse loci. The similar use of D and J segments suggests that the recognition of recombination signal sequences is all but identical in mouse and man. The insertion of N nucleotides in the H chains, but not the L chains, indicates that the transloci rearrange at the same time as their endogenous counterparts. The major limitation on the potential variability of the repertoire arises from the limited number of V region gene segments present in the HuIgHeavy YAC, apparent by the overexpression of the VH1-2 segment. The Ig rearrangement process is affected by variations both in the promoter of the V region (40) and in the recombination signal sequences (41). Given the small number of V regions in the locus, even a relatively small advantage in rearrangement frequency during B cell development may manifest as a large bias in the utilization of the region in the Ig repertoire (42). The sequencing of the human H chain locus indicates that the segments in the HuIgH YAC differ in both the promoter region and the recombination signals (43), which may account for the observed utilization of VH1-2 in preference to the other segments. The inclusion of more V region segments in the translocus should overcome this limitation.

The presence of point mutations in rearranged Ig genes indicates that they are from a B cell that has undergone affinity maturation following exposure to Ag. Although it has been reported that mutations can be present in the H or L chain but not in the partner sequence, the most common state is that both chains contain mutations (44, 45). Here we found frequent mutations in L chain rearrangements, but almost no mutation in the H chains. The contrast was particularly evident in the paired sequences isolated from hybridoma clones. The presence of mutations in the L chain rearrangements confirms that the B cells expressing human Ab are able to participate in germinal center reactions, and the biased distribution of replacement mutations indicates that these mutations were subject to antigenic selection (34). The consistent absence of mutations in the H chain rearrangements, with only four of 41 rearrangements containing two or more point mutations, would indicate that the H chain translocus provides a poor target for the introduction of somatic hypermutation mechanism. Efficient hypermutation of the VH segments has been described in the case of several, but not all, human IgH transloci (6, 10, 13, 46). This variation could indicate a sensitivity to integration position effects, particularly since transloci are unlikely to contain the full complement of IgH locus cis-acting regulatory elements.

We have produced mice carrying germline H chain and κ and λ L chain transloci in a background where the mouse H chain and κ L chain are nonfunctional. The transloci can substitute for their mouse counterparts, leading to the formation of diverse repertoires of fully human Ab. Sequence analysis indicates that the transloci are rearranged at the same stage as their mouse equivalents, and that the segments are used as they would be in human cells. The mice we have developed will be useful in the study of Ig gene rearrangement during B cell development. They can be used to produce fully human mAb for therapy, avoiding the adverse reactions that can be induced by rodent Ab.

We thank Dr. J. White for assistance with Southern blotting, and Nigel Miller for FACS sorting. We are grateful to Drs. D. Kitamura and K. Rajewsky for providing the μMT mice, and to Dr. D. Melton for kindly providing the HM-1 ES cells.

1

This work was supported in part by the Babraham Institute through a Biotechnology and Biological Sciences Research Council Competitive Strategic Grant, and by a European Union Biotechnology Project Grant. I.C.N. holds a Howard Florey Fellowship from the Royal Society.

5

Abbreviations used in this paper: YAC, yeast artificial chromosome; H chain, heavy chain of Ig; IGF, insulin-like growth factor; L chain, light chain of Ig; PLAP, placental alkaline phosphatase. (The term “four-feature” mice refers to animals carrying the HuIgHeavy and HuIgκ YACs and the μMT−/− and Moκ−/− knockouts. Similarly, the term “five-feature” mice refers to mice with the HuIgHeavy, HuIgκ, and HuIgλ YACs with the μMT−/− and Moκ−/− knockouts.)

1
Kuzel, T. M., S.T. Rosen.
1994
. Antibodies in the treatment of human cancer.
Curr. Opin. Oncol.
6
:
622
2
Khazaeli, M. B., R. M. Conry, A. F. LoBuglio.
1994
. Human immune response to monoclonal antibodies.
J. Immunother.
15
:
42
3
Griffiths, A. D., S. C. Williams, O. Hartley, I. M. Tomlinson, P. Waterhouse, W. L. Crosby, R. E. Kontermann, P. T. Jones, N. M. Low, T. J. Allison, et al
1994
. Isolation of high affinity human antibodies directly from large synthetic repertoires.
EMBO J.
13
:
3245
4
Brüggemann, M., H. M. Caskey, C. Teale, H. Waldmann, G. T. Williams, M. A. Surani, M. S. Neuberger.
1989
. A repertoire of monoclonal antibodies with human heavy chains from transgenic mice.
Proc. Natl. Acad. Sci. USA
86
:
6709
5
Brüggemann, M., C. Spicer, L. Buluwela, I. Rosewell, S. Barton, M. A. Surani, T. H. Rabbitts.
1991
. Human antibody production in transgenic mice: expression from 100 kb of the human IgH locus.
Eur. J. Immunol.
21
:
1323
6
Taylor, L. D., C. E. Carmack, D. Huszar, K. M. Higgins, R. Mashayekh, G. Sequar, S. R. Schramm, C. C. Kuo, S. L. O’Donnell, R. M. Kay, et al
1994
. Human immunoglobulin transgenes undergo rearrangement, somatic mutation and class switching in mice that lack endogenous IgM.
Int. Immunol.
6
:
579
7
Harmer, I. J., R. A. Mageed, A. Kaminski, P. Charles, M. Brüggemann, C. G. Mackworth-Young.
1996
. Chimaeric monoclonal antibodies encoded by the human VH26 gene from naive transgenic mice display a wide range of antigen-binding specificities.
Immunology
88
:
174
8
Davies, N. P., I. R. Rosewell, M. Brüggemann.
1992
. Targeted alterations in yeast artificial chromosomes for inter-species gene transfer.
Nucleic Acids Res.
20
:
2693
9
Zou, X., J. Xian, N. P. Davies, A. V. Popov, M. Brüggemann.
1996
. Dominant expression of a 1.3 Mb human Igκ locus replacing mouse light chain production.
FASEB J.
10
:
1227
10
Mendez, M. J., L. L. Green, J. R. Corvalan, X. C. Jia, C. E. Maynard-Currie, X. D. Yang, M. L. Gallo, D. M. Louie, D. V. Lee, K. L. Erickson, et al
1997
. Functional transplant of megabase human immunoglobulin loci recapitulates human antibody response in mice.
Nat. Genet.
15
:
146
11
Green, L. L., M. C. Hardy, C. E. Maynard-Currie, H. Tsuda, D. M. Louie, M. J. Mendez, H. Abderrahim, M. Noguchi, D. H. Smith, Y. Zeng, et al
1994
. Antigen-specific human monoclonal antibodies from mice engineered with human Ig heavy and light chain YACs.
Nat. Genet.
7
:
13
12
Brüggemann, M., M. S. Neuberger.
1996
. Strategies for expressing human antibody repertoires in transgenic mice.
Immunol. Today
17
:
391
13
Wagner, S. D., G. Gross, G. P. Cook, S. L. Davies, M. S. Neuberger.
1996
. Antibody expression from the core region of the human IgH locus reconstructed in transgenic mice using bacteriophage P1 clones.
Genomics
35
:
405
14
Popov, A. V., X. Zou, J. Xian, I. C. Nicholson, M. Brüggemann.
1999
. A human Igλ locus is similarly well expressed in mice and humans.
J. Exp. Med.
189
:
1611
15
Davies, N. P., A. V. Popov, X. Zou, M. Brüggemann.
1996
. Human antibody repertoires in transgenic mice: Manipulation and transfer of YACs. J. McCafferty, and H. R. Hoogenboom, and D. J. Chiswell, eds.
Antibody Engineering: A Practical Approach
59
-76. IRL, Oxford.
16
Selfridge, J., A. M. Pow, J. McWhir, T. M. Magin, D. W. Melton.
1992
. Gene targeting using a mouse HPRT minigene/HPRT-deficient embryonic stem cell system: inactivation of the mouse ERCC-1 gene.
Somatic Cell Mol. Genet.
18
:
325
17
Hogan, B., R. Beddington, F. Costantini, E. Lacy.
1994
.
Manipulating the Mouse Embryo: A Laboratory Manual
Cold Spring Harbor Press, Cold Spring Harbor.
18
Kitamura, D., J. Roes, R. Kühn, K. Rajewsky.
1991
. A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin μ chain gene.
Nature
350
:
423
19
Zou, X., J. Xian, A. V. Popov, I. R. Rosewell, M. Müller, M. Brüggemann.
1995
. Subtle differences in antibody responses and hypermutation of λ light chains in mice with a disrupted κ constant region.
Eur. J. Immunol.
25
:
2154
20
Klein, U., G. Klein, B. Ehlin-Henriksson, K. Rajewsky, R. Küppers.
1995
. Burkitt’s lymphoma is a malignancy of mature B cells expressing somatically mutated V region genes.
Mol. Med.
1
:
495
21
Xian, J., X. Zou, A. V. Popov, C. Mundt, N. Miller, G. T. Williams, S. L. Davies, M. S. Neuberger, M. Brüggemann.
1998
. Comparison of the performance of a plasmid-based human Igκ minilocus and YAC-based human Igκ transloci for the production of human antibody repertoires in transgenic mice.
Transgenics
2
:
333
22
Lefranc, M. P., V. Giudicelli, C. Ginestoux, J. Bodmer, W. Müller, R. Bontrop, M. Lemaitre, A. Malik, V. Barbie, D. Chaume.
1999
. IMGT, the international ImMunoGeneTics database.
Nucleic Acids Res.
27
:
209
23
Pech, M., H. Smola, H. D. Pohlenz, B. Straubinger, R. Gerl, H. G. Zachau.
1985
. A large section of the gene locus encoding human immunoglobulin variable regions of the kappa type is duplicated.
J. Mol. Biol.
183
:
291
24
Popov, A. V., C. Bützler, J. P. Frippiat, M. P. Lefranc, M. Brüggemann.
1996
. Assembly and extension of yeast artificial chromosomes to build up a large locus.
Gene
177
:
195
25
Asenbauer, H., G. Combriato, H. G. Klobeck.
1999
. The immunoglobulin λ light chain enhancer consists of three modules which synergize in activation of transcription.
Eur. J. Immunol.
29
:
713
26
Haury, M., A. Sundblad, A. Grandien, C. Barreau, A. Coutinho, A. Nobrega.
1997
. The repertoire of serum IgM in normal mice is largely independent of external antigenic contact.
Eur. J. Immunol.
27
:
1557
27
Wagner, S. D., G. T. Williams, T. Larson, M. S. Neuberger, D. Kitamura, K. Rajewsky, J. Xian, M. Brüggemann.
1994
. Antibodies generated from human immunoglobulin miniloci in transgenic mice.
Nucleic Acids Res.
22
:
1389
28
Brezinschek, H. P., R. I. Brezinschek, P. E. Lipsky.
1995
. Analysis of the heavy chain repertoire of human peripheral B cells using single-cell polymerase chain reaction.
J. Immunol.
155
:
190
29
Ichihara, Y., H. Hayashida, S. Miyazawa, Y. Kurosawa.
1989
. Only DFL16, DSP2, and DQ52 gene families exist in mouse immunoglobulin heavy chain diversity gene loci, of which DFL16 and DSP2 originate from the same primordial DH gene.
Eur. J. Immunol.
19
:
1849
30
Corbett, S. J., I. M. Tomlinson, E. L. L. Sonnhammer, D. Buck, G. Winter.
1997
. Sequence of the human immunoglobulin diversity (D) segment locus: a systematic analysis provides no evidence for the use of DIR segments, inverted D segments, “minor” D segments or D-D recombination.
J. Mol. Biol.
270
:
587
31
Kantor, A. B., C. E. Merrill, L. A. Herzenberg, J. L. Hillson.
1997
. An unbiased analysis of V(H)-D-J(H) sequences from B-1a, B-1b, and conventional B cells.
J. Immunol.
158
:
1175
32
Victor, K. D., K. Vu, A. J. Feeney.
1994
. Limited junctional diversity in κ light chains: junctional sequences from CD43+B220+ early B cell progenitors resemble those from peripheral B cells.
J. Immunol.
152
:
3467
33
Feeney, A. J., G. Lugo, G. Escuro.
1997
. Human cord blood κ repertoire.
J. Immunol.
158
:
3761
34
Weiss, U., K. Rajewsky.
1990
. The repertoire of somatic antibody mutants accumulating in the memory compartment after primary immunization is restricted through affinity maturation and mirrors that expressed in the secondary response.
J. Exp. Med.
172
:
1681
35
Jakobovits, A., G. J. Vergara, J. L. Kennedy, J. F. Hales, R. P. McGuinness, D. E. Casentini-Borocz, D. G. Brenner, G. R. Otten.
1993
. Analysis of homozygous mutant chimeric mice: deletion of the immunoglobulin heavy-chain joining region blocks B-cell development and antibody production.
Proc. Natl. Acad. Sci. USA
90
:
2551
36
Huber, C., E. Huber, A. Lautner-Rieske, K. F. Schäble, H. G. Zachau.
1993
. The human immunoglobulin κ locus: characterization of the partially duplicated L regions.
Eur. J. Immunol.
23
:
2860
37
Frippiat, J. P., S. C. Williams, I. M. Tomlinson, G. P. Cook, D. Cherif, D. Le Paslier, J. E. Collins, I. Dunham, G. Winter, M. P. Lefranc.
1995
. Organization of the human immunoglobulin lambda light-chain locus on chromosome 22q11.2.
Hum. Mol. Genet.
4
:
983
38
Cox, J. P., I. M. Tomlinson, G. Winter.
1994
. A directory of human germ-line Vκ segments reveals a strong bias in their usage.
Eur. J. Immunol.
24
:
827
39
Ignatovich, O., I. M. Tomlinson, P. T. Jones, G. Winter.
1997
. The creation of diversity in the human immunoglobulin Vλ repertoire.
J. Mol. Biol.
268
:
69
40
Stiernholm, N. B., N. L. Berinstein.
1995
. A mutated promoter of a human Ig Vλ gene segment is associated with reduced germ-line transcription and a low frequency of rearrangement.
J. Immunol.
154
:
1748
41
Nadel, B., A. Tang, G. Escuro, G. Lugo, A. J. Feeney.
1998
. Sequence of the spacer in the recombination signal sequence affects V(D)J rearrangement frequency and correlates with nonrandom Vκ usage in vivo.
J. Exp. Med.
187
:
1495
42
Yu, C. C., M. Larijani, I. N. Miljanic, G. E. Wu.
1998
. Differential usage of VH gene segments is mediated by cis elements.
J. Immunol.
161
:
3444
43
Matsuda, F., K. Ishii, P. Bourvagnet, K. Kuma, H. Hayashida, T. Miyata, T. Honjo.
1998
. The complete nucleotide sequence of the human immunoglobulin heavy chain variable region locus.
J. Exp. Med.
188
:
2151
44
Brezinschek, H. P., S. J. Foster, T. Dörner, R. I. Brezinschek, P. E. Lipsky.
1998
. Pairing of variable heavy and variable κ chains in individual naive and memory B cells.
J. Immunol.
160
:
4762
45
de Wildt, R. M. T., R. M. A. Hoet, W. J. van Venrooij, I. M. Tomlinson, G. Winter.
1999
. Analysis of heavy and light chain pairings indicates that receptor editing shapes the human antibody repertoire.
J. Mol. Biol.
285
:
895
46
Wagner, S. D., A. V. Popov, S. L. Davies, J. Xian, M. S. Neuberger, M. Brüggemann.
1994
. The diversity of antigen-specific monoclonal antibodies from transgenic mice bearing human immunoglobulin gene miniloci.
Eur. J. Immunol.
24
:
2672
47
Kabat, E. A., T.T. Wu, M. Reid-Miller, K. S. Gottesman.
1987
.
Sequences of Proteins of Immunological Interest
U.S. Government Printing Office, Bethesda.