PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Immunogenetics. Author manuscript; available in PMC 2010 November 2.
Published in final edited form as:
PMCID: PMC2970566
NIHMSID: NIHMS244980

Vκ polymorphisms in NOD mice are spread throughout the entire immunoglobulin kappa locus and are shared by other autoimmune strains

Abstract

The diversity of immunoglobulin (Ig) and T cell receptor (TCR) genes available to form the lymphocyte repertoire has the capacity to produce a broad array of both protective and harmful specificities. In type 1 diabetes (T1D), the presence of antibodies to insulin and other islet antigens predicts disease development in both mice and humans, and demonstrate that immune tolerance is lost early in the disease process. Anti-insulin T cells isolated from T1D-prone non-obese diabetic (NOD) mice use polymorphic TCRα chains, suggesting that the available T cell repertoire is altered in these autoimmune mice. To probe whether insulin-binding B cells also possess polymorphic V genes, Ig light chains were isolated and sequenced from NOD mice that harbor an Ig heavy chain transgene. Three insulin-binding Vκ genes were identified, all of which were polymorphic to the closest germline sequence matches present in the GenBank database. Additional analysis of over 300 light chain sequences from multiple sources, including germline DNA, shows that polymorphisms are spread throughout the entire NOD Igκ locus, as these polymorphic sequences represent 43 distinct Vκ genes which belong to 14 Vκ families. Database searches reveal that a majority of polymorphic Vκ genes identified in NOD are identical to Vκ genes isolated from SLE-prone NZBxNZW F1 or MRL strains of mice, suggesting that a shared Igκ haplotype may be present. Predicted amino acid changes preferentially occur in CDR, and thus could alter antigen recognition by the germline B cell repertoire of autoimmune versus non-autoimmune mouse strains.

Keywords: B cell, Type 1 diabetes, Autoimmunity, Immunoglobulin kappa, Light chain

Introduction

Autoimmunity arises from a breach in tolerance in one or more branches of the immune system, which results in the production and survival of autoreactive cells. Autoimmune damage caused by these cells manifests in a diverse array of autoimmune diseases which afflict 5–8% of the population (The Autoimmune Diseases Coordinating Committee and National Institutes of Health. Progress in Autoimmune Diseases Research 2005). Autoantibodies are present in the prodrome of many autoimmune diseases, including type 1 diabetes (T1D), in which the presence of islet-reactive autoantibodies are predictive of disease development in both humans and mouse models of disease (Yu et al. 2000). These autoantibodies provide direct evidence that B cell tolerance is interrupted in T1D, as effective tolerance should prevent autoreactive B cells from differentiating into the effector plasma B cell stage.

The antigen specificity of a B cell is determined by rearrangement of immunoglobulin gene segments to produce functionally expressed heavy and light chains, which pair to generate a B cell receptor (BCR). Factors which could contribute to the formation of autoreactive B lymphocytes therefore may include differences in the germline repertoire of these gene segments, as well as ineffective removal of autoreactive clones from the mature repertoire. Antigen specificity of T and B cells has been shown to be important for promoting T1D development (Hulbert et al. 2001; Lennon et al. 2009; Silveira et al. 2002). Interestingly, Simone, et al. has found that the T cell receptor α (TCR ) repertoire expressed by anti-insulin T cells is highly restricted and polymorphic in the non-obese diabetic (NOD) mouse (Simone et al. 1997a, b). As TCRα and BCR light chains are structurally similar, we were therefore interested in identifying whether the NOD BCR light chain locus may also be polymorphic. Furthermore, understanding the relationship between germline sequences and V region structures expressed by autoantibodies has important consequences for assessing affinity maturation and T cell dependence.

The studies presented here demonstrate that insulin-binding light chains isolated from NOD mice are polymorphic to germline Vκ gene sequences available in the IgBLAST database that is referenced to genome sequencing of C57BL/6 and other common non-autoimmune strains. Polymorphisms do not appear to be restricted to insulin-binding Vκ, as sequencing of over 300 Vκ clones isolated from various organs as well as germline DNA in NOD mice revealed 43 distinct NOD Vκ genes, nearly all of which are polymorphic to reference germline sequences. In striking contrast, the majority of NOD Vκ genes were found to be 100% identical to one or more light chains found to be expressed by systemic lupus erythematosus (SLE)-prone MRL/lpr or NZBxNZW F1 mouse strains. The nt differences in the NOD strain encode aa changes that are observed throughout FWR and CDR, but which are preferentially found in the CDR. We therefore propose that portions of the Igκ haplotype of the NOD mouse may be shared among several autoimmune disease-prone strains of mice, with the potential consequence that alterations in the antigen specificity of the germline repertoire could exist.

Materials and methods

Animals

Conventional (wild-type) NOD mice used in these studies were originally obtained from Taconic Laboratories. The anti-insulin VH125Tg [Cg-Tg(Igh-6/Igh-V125)2Jwt/JwtJ], The Jackson Laboratory) heavy chain transgenic mice used in this study harbor a non-targeted VH transgene on NOD backgrounds as described previously (Hulbert et al. 2001; Rojas et al. 2001). All data are derived from lines that have been backcrossed >20 generations to NOD, and which are hemizygous for all transgenes indicated. All mice were housed under sterile housing conditions, and all studies were approved by the Institutional Animal Care and Use Committee of Vanderbilt University.

Hybridoma generation

VH125Tg NOD mice were immunized intraperitoneally with 0.6 mg of porcine insulin covalently coupled to Brucella abortus ring test antigen (obtained from the US Department of Agriculture, Ames, IA), 100 µL injection volume per animal. The coupling procedure was carried out using insulin acylated with m-maleimidobenzoyl-N-hydrox-ysuccinimide ester (Pierce Chemical Co.) and B. abortus thiolated with methyl-4-mercaptobutyrimidate (Pierce). Splenocytes were harvested after 3 d and fused using polyethyleneglycol to myeloma cells to generate hybridomas (Thomas et al. 1997, 2002). This approach has uniformly identified germline-encoded antibodies. To detect the production of anti-insulin antibody, 96-well microtiter plates (Immulon II) were coated with 1 µg/mL human insulin in 1× borate buffered saline overnight at 37°C. Plates were washed and hybridoma supernatants were applied and incubated 1 h at room temperature. For competitive inhibition of binding, 50 µg/mL soluble insulin was incubated with hybridoma supernatants. Isotype-specific goat anti-mouse IgM or IgG conjugated to alkaline phosphatase (Southern Biotech) was used for detection. Anti-insulin clone 1F11 (mAb301) was used as a positive control for IgM (Thomas et al. 1997) and mAb125 was used as a control for IgG (Thomas and Hulbert 1996) anti-insulin antibody. All dilutions and washes were conducted at room temperature using 1× phosphate buffered saline (PBS) plus 0.2% Tween 20. Results are reported as mean OD405 for using paranitrophenyl phosphate (Sigma-Aldrich) as substrate and an automated ELISA reader. Hybridomas producing anti-insulin antibody were cloned by limiting dilution and expanded. ELISA was used to confirm insulin-binding antibody (as above), and RNA was purified from expanded clones and Vκ genes were amplified and sequenced as described below.

Cell isolation and flow cytometry

Bone marrow was eluted from the long bones of VH125Tg NOD mice with HBSS (Invitrogen) plus 10% FBS (HyClone). Red blood cells (RBC) were lysed and cells were resuspended at 2×106 cells/ml in complete bone marrow culture medium [DMEM plus 10% FBS (HyClone) plus l-glutamine plus HEPES buffer plus MEM sodium pyruvate plus gentamicin plus 2×10−5M 2-ME (Invitrogen) plus 15 ng/ml human rIL-7 (PeproTech)] and cultured for 5 days in a 37°C CO2 incubator to enrich for B cells. Antibody reagents reactive with B220 (6B2), IgMa (DS-1), 7-aminoactinomycin D (7AAD), or 4', 6-diamidino-2-phenylindole (DAPI) were used for flow cytometry (BD Pharmingen). Insulin (Sigma) was biotinylated at pH 8.0 in bicine buffer using biotin N-hydroxysuccinimide ester (Sigma) and detected with fluorochrome-labeled streptavidin (BD Pharmingen). RNAwas subsequently isolated from insulin-binding B cells sorted through flow cytometry with a BD FACSAria cell sorter. Spleens were harvested and red blood cells were lysed, and splenocytes were resuspended. RNA was directly purified from splenocytes, or following additional purification (flow cytometry sorting of insulin-binding B cells, or separation of B cells bound to plate-bound insulin, as described previously by Woodward and Thomas 2005). Pancreata and pancreatic lymph nodes were harvested as previously described (Kendall et al. 2007) and placed directly into RNAlater (Ambion). For islet isolation, pancreata were macerated in 1 ml of HBSS. Tissue was agitated with 3 mg/mL collagenase P in HBSS for 12 min at 37°C. Pancreatic material was washed and islets were hand-picked using a dissecting microscope. Islets were placed in culture overnight at 37°C with 5% CO2, and islets with extruded lymphocytes were selected and placed into RNAlater (Kendall et al. 2007). Germinal center B lymphocytes from pancreatic lymph nodes were sorted with a FACSAria Cell Sorting System (BD Biosciences), gating on GL7+ B220+ 7-aminoactinomycin D (7AAD) lymphocytes. WinMDI 2.8 software (Dr. J. Trotter, Scripps Institute, San Diego, CA) was used for analysis. RNA extraction was performed with the RNAqueous-Midi or Micro Kit (Ambion), following the manufacturer’s instructions, or using Tri Reagent (Molecular Research Center) for some of the spleen samples.

Determination of Vκ usage

First-strand cDNA was generated from total RNA using Superscript II RT (Invitrogen) and 0.67 µg oligo-dT primer (Amersham Biosciences) in a standard protocol. Vκ sequences were amplified from first-strand cDNA using the following primers: murine Cκ primer—5′GGA TAC AGT TGG TGC AGC ATC and murine VκA—5′ATT GTK MTS ACM CAR TCT CCA, VκB, 5′-GAT RTT KTG RTR ACB CAR RM, or VκC, 5′-AYA TYN WGM TGA CHC ARW CTM M, where K = G or T, M = A or C, S = C or G, R = A or G. Individual clones are each derived from independent pools of RNA. Vκ sequences were amplified, cloned, and sequenced as described previously (Woodward and Thomas 2005). Vκ gene segment sequence alignments (excluding sequence provided by the degenerate 5′ primer) were assigned using the ImMunoGeneTics (IMGT) database (www.imgt.cines.fr:8104/) and IgBLAST (http://www.ncbi.nlm.nih.gov/igblast/). Specific alignments are provided with the genomic C57BL/6 sequence, build 37.1 as reference. Accession numbers for representative sequences from each of the NOD Vκ genes are provided in Table 2. Sequence alignments of clones included in this study are provided as online supplemental data.

Table 2
Location of NOD polymorphisms and GenBank accession numbers

Germline Vκ genes from NOD tail DNA samples were amplified using the following temperatures in PCR: 94°C 1 min, 55, 57, or 60°C 1 min, 72°C 1 min, with the following primers for 40 cycles: Vκ1 GL FWD: TTT GCA TAT TGC TCC CTA GGG A, Vκ1 GL REV: TTG TTA GGG TCT GTATCA CTG TGG GA or Vκ4 GL FWD: TAT TTG CAT ATT TCA TTT TCA GTA ACC ACA, Vκ4 GL REV: TGT TCC AGT CTG TAT CAC TGT GTG; FWD primers recognized the conserved octamer regulatory sequence, REV primers recognized the heptamer and a portion of the nonamer of the RSS. Variation of the annealing temperature aided in amplification of different Vκ genes. Germline genes were cloned, sequenced, and identified as above. Accession numbers for germline sequences deposited into GenBank are HM444791-HM444810.

Results

Insulin-binding NOD Vκ genes are polymorphic to reference germline Vκ genes

VH125Tg NOD mice, which harbor an anti-insulin heavy chain transgene, show an increased percentage of insulin-binding spleen B cells when compared to VH125Tg C57BL/6 mice (Hulbert et al. 2001; Woodward and Thomas 2005). Differences in the endogenous light chain repertoire between the two strains, as is found among insulin-binding, structurally homologous TCRα chains (Simone et al. 1997a, b), could help explain this difference. VH125Tg NOD mice were therefore immunized with insulin conjugated to B. abortus ring test antigen and hybridomas were generated from splenocytes harvested on d3. Hybridomas were cloned by limiting dilution and insulin-binding specificity was confirmed by screening the supernatants of the clones by ELISA (Fig. 1a). Three different insulin-binding Vκ light chain sequences were identified from multiple hybridomas by RT-PCR. These sequences showed conserved nt changes when compared to germline sequences present in the IgBLAST and IMGT databases. For consistency, sequences were compared to the C57BL/6 reference strain, for which complete genome sequencing is available. Each of these light chains is observed to differ at multiple nucleotides within both framework (FWR) and complementarity determining regions (CDR), boundaries for which are designated by the Kabat system (Fig. 1b – d), when compared to counter-part C57BL/6 Vκ genes 3–4, 4–57-1, and 4–74. Insulin-B. abortus activates B cells independent of T cell help; thus these finding suggest that either anti-insulin NOD Vκ genes are polymorphic to C57BL/6 Vκ genes, or that the immunization procedure captured mutated Vκ genes. The term “polymorphism” is defined as follows: “genetic polymorphism is variability at a gene locus in which the variants occur at a frequency of greater than 1%” (Janeway et al. 2001), and refers to the observed nt changes present in the NOD strain compared with the C57BL/6 reference strain, many of which encode aa changes.

Fig. 1
Insulin-binding Vκ genes are polymorphic to C57BL/6 germline sequences. Hybridomas generated from VH125Tg NOD splenic B cells were cloned at limiting dilution. a Hybridoma clone supernatants were tested by ELISA to identify insulin-binding antibody ...

Polymorphisms are observed in NOD Vκ genes throughout the Igκ locus

To determine the origin of nt differences observed in the anti-insulin NOD Vκ genes isolated from insulin-specific hybridomas, RNA was isolated from B cells harvested from multiple organs of non-immunized NOD mice, including the bone marrow, spleen, lymph nodes, and pancreas and Vκ genes were amplified and sequenced. Vκ sequences included in this analysis are those for which three or more independent sequences show 100% identity, either with other sequences determined in these studies or with sequences published in the 2009 GenBank database. Using these criteria, 43 unique Vκ genes were identified which belong to 14 different Vκ families based on the IMGT classification system (Table 1). Over 300 light chain sequences were isolated from various organs of non-immunized mice, as well as germline DNA; these studies thus provide the widest coverage of NOD Vκ gene sequencing published to date. For ease, the closest reference C57BL/6 Vκ gene match from the IMGT database for each NOD Vκ gene, as well as the Zachau system designation are listed in Table 1.

Table 1
Light chain polymorphisms are present in the majority of NOD Vκ genes, which are shared among other autoimmune strains of mice

The nt changes observed among the three insulin-binding hybridoma Vκ genes depicted in Fig. 1 are shared among corresponding Vκ clones derived from multiple organs isolated directly from non-immunized mice (SD Figs. 14, 19 and 24). Furthermore, nt changes are present in 93% of the NOD Vκ genes identified (Table 1). The NOD Vκ genes presented in Table 1 range from 91.2% to rarely 100% nt identity with the closest C57BL/6 germline matches (SD Figs. 1–43). The number of nt differences present range from 0 to 24, with an average of 8 nt differences per Vκ gene (Fig. 2 and Table 2). The majority of the observed nt polymorphisms predict amino acid replacement, however silent changes are also predicted (Fig. 2). There were four pairs of NOD Vκ genes which were distinct from each other, but which showed the highest sequence identity with the same reference germline Vκ genes (Vκ1–133, Vκ1–135, Vκ3–4, and Vκ4–57-1), thus complicating assignment of independent Vκ gene identities. The closest C57BL/6 match is thus insufficient to formally assign NOD Vκ gene identities; rather, this sequence information will need to be combined with Igκ locus position information derived from sequencing the entire NOD Igκ locus to make individual gene assignments.

Fig. 2
Polymorphisms are present in the majority of identified NOD Vκ genes and are spread throughout the Igκ locus. NOD Vκ genes for which three or more sequences that differed from each other by ≤2 nt changes are included in ...

Despite originating from multiple different organs, it is possible that the nt changes observed in the NOD Vκ gene sequences arise from somatic hypermutation (SHM), perhaps due to specific autoantigen-driven selection. However, nearly 70% of the identified NOD Vκ genes show 100% sequence identity with other published Vκ gene sequences identified from other autoimmune mouse strains, particularly NZBxNZW F1 and MRL/lpr mouse strains. As SLE results from targeting of a different cohort of autoantigens, it is improbable that all of the nt changes observed are due to SHM. Furthermore, the nt changes observed are also found in a panel of NOD Vκ gene sequences produced by another laboratory, confirming that they are not specific to our colony of NOD mice (Table 1, Carrillo et al. 2008). To confirm that the nt changes observed were germline-encoded, and not due to SHM, light chains were amplified and cloned from NOD genomic tail DNA. The same nt changes that were observed in sequences isolated from B cells in various organs of the NOD mouse were present in 15/15 corresponding unique germline sequences retrieved from tail DNA (Table 2 and SD Figs. 2– 11, 18–19, 21, 23, 25–26, and 33–34), confirming that the observed nt changes are germline-encoded polymorphisms. These data suggest that the majority of NOD Vκ genes are polymorphic to the reference database of Vκ genes compiled from C57BL/6 and other non-autoimmune strains of mice, many of which are identical to Vκ sequences isolated from other autoimmune disease-prone NZBxNZW F1 and MRL strains of mice.

The nucleotide polymorphisms in NOD Vκ genes are found in FWR, but occur most frequently in CDR

CDR are responsible for antigen-binding specificity of the antibody molecule, thus we were interested in identifying whether the NOD Vκ nt differences occurred in these antigen-contacting regions. The location of nt changes were categorized as occurring in either FWR1, CDR1, FWR2, CDR2, FWR3, or CDR3, based on the Kabat system-defined boundaries for each germline Vκ gene (Table 2). To analyze whether the nt differences were preferentially located in the FWR or CDR portions of the antibody molecule, the per site rate of nt changes was calculated. For example, the total number of nt changes in FWR1 was divided by the total number of nt in FWR1, and the average value observed for all 43 Vκ genes was calculated to produce the average frequency of nt changes in FWR1. This was repeated for FWR2, FWR3, CDR1, CDR2, and CDR3. The average frequency of changes in FWR1 was divided by the sum of the average frequency of changes in all of the FWR and CDR to calculate the relative proportion of nt changes that were occurring in FWR1, which was repeated for the other regions. As shown in Fig. 3, the relative proportion of nt polymorphisms found in the CDR (69%) was greater than that found in the FWR (31%), and was distributed among each of the CDR1–3 without substantial bias. These data therefore show that nt differences are present in both the FWR and CDR, but are preferentially targeted to the CDR.

Fig. 3
The frequency of NOD Vκ nucleotide changes is increased in CDR regions. The per site rate of nt changes in each FWR and CDR was calculated, e.g., the total number of nt changes in FWR1 was divided by the total number of nt in FWR1 for each of ...

Amino acid replacements are found throughout the FWR and CDR, but occur most frequently in CDR

To identify whether the nt polymorphisms observed in the NOD Vκ genes resulted in changes in the protein sequence, aa alignments of NOD and reference Vκ genes were compared. The aa sequences of NOD Vκ genes presented are predicted translations of experimentally determined nt sequences. In this analysis, replacement (R) or silent (S) changes were enumerated for each gene segment (Table 3) within Kabat-defined FWR and CDR boundaries. The number of amino acid changes per NOD Vκ gene is shown in Fig. 4. The sum of R and S aa changes encoded by nt was calculated for all 43 Vκ genes analyzed and these totals were divided to yield an average R/S for each FWR and CDR. The average R/S was as follows: FWR1: 0.8, FWR2: 1.9, FWR3: 1.0, CDR1: 3.9, CDR2: 3.2, CDR3: 12.8; the average R/S for FWR=1.3, CDR=6.6; Table 4). Further analysis of conservative (e.g., T>S) and non-conservative (e.g., E>K) aa changes was performed as follows: the sum of the total number of conservative and non-conservative aa changes for each NOD Vκ gene in each FWR or CDR was calculated, as well as the percentage of non-conservative aa changes (number of non-conservative aa changes divided by the total # of aa changes) for each region, and these data are shown in Table 4. These data show that many of the aa replacements are non-conservative, that they occur more frequently in CDR, and that they are particularly concentrated in CDR3.

Fig. 4
Nucleotide polymorphisms result in amino acid changes in NOD Vκ genes. NOD Vκ aa sequences are based on predicted translations from NOD nt sequences. The number of aa changes in each NOD Vκ gene is shown, along with the putative ...
Table 3
Location of NOD Vκ nt polymorphisms which result in aa replacements (R), or are silent (S) in FWR and CDR
Table 4
Total number of replaced (R) and silent (S), as well as conservative and non-conservative amino acid changes present in FWR and CDR in NOD Vκ genes, based on predicted amino acid sequences

Discussion

The findings presented here suggest that many, and likely the large majority of NOD Vκ genes are polymorphic to Vκ genes in the reference database of common mouse strains and the sequenced genome of C57BL/6 mice. The sequences published here are believed to be germline, and not the result of SHM, as they are identical among clones isolated from independent animals and B cells derived from multiple NOD organs. Among the 43 NOD Vκ genes analyzed, 15/15 unique corresponding sequences amplified from germline DNA confirm 100% of the polymorphisms observed in the other sequences (Table 2 and SD Figs. 2–7, 11, 18–19, 21, 23, 25–26, and 33–34). Furthermore they share identity with sequences found in other autoimmune disease-prone strains of mice such as MRL and NZBxNZW F1, which target a different cadre of autoantigens. Polymorphisms are observed in all of the different Vκ families represented in this study, and are spread throughout the Igκ locus (Table 1, Figs. 2 and and5).5). The NOD mouse therefore possesses a unique Igκ locus that is polymorphic to Igκ reference sequences derived primarily from non-autoimmune strains, such as C57BL/6; however, portions are shared with SLE-prone mouse strains. It is possible that the altered available germline Vκ genes may shape the repertoire of antigen specificities generated in the developing B cell repertoire, as the per site rate of nt changes and the ratio of R/S is increased for the CDR (Fig. 3 and Table 4). These polymorphisms may thus contribute to B cell autoimmunity in the NOD mouse; however, whether or not these changes enhance repertoire autoreactivity requires further investigation.

Fig. 5
NOD Vκ sequence matches with NZB x NZW F1 Vκ cluster to the 5′ region, whereas matches with MRL Vκ cluster to the 3′ region of the Igκ locus. The closest C57BL/6 germline match was identified by IgBLAST, ...

The presence of polymorphisms in the Igκ locus have been previously reported in a widespread analysis of multiple strains (Kofler et al. 1989), however this is the first study to demonstrate the broad array of polymorphisms present in the NOD mouse. Many previously published Vκ gene sequences derived from autoimmune strains were thought to be derived from SHM. In light of our findings that many of these “mutated” sequences show 100% identity with other autoimmune Vκ sequences, we propose that in many cases, these differences arise from germline polymorphisms, rather than from SHM. It is unlikely that B cells isolated from different autoimmune mice share the same autoantigen specificity, as T1D and SLE arise from pathogenic targeting of a different cohort of autoantigens. Given the likely diversity in the antigens responsible for driving selection, it is improbable that SHM would produce identical nt changes in both T1D and SLE-prone strains of mice.

The IMGT and Kabat gene assignments given to particular NOD Vκ genes are provided throughout figures and tables to give a helpful frame of reference. These assignments cannot yet be confirmed however, as NOD Igκ locus has not yet been sequenced, and thus the position and therefore identity of individual NOD Vκ genes needs to be verified. Many germline Vκ genes do not greatly differ in nt sequence from one another, particularly among the Vκ4 family. Eight NOD Vκ genes identified in this study (NOD Vκ gene 6, 7, 8, 9, 13, 14, 19, 20, Table 1) showed the closest sequence identity to the same 4 reference C57BL/6 germline genes, despite clearly being distinct sequences from one another. For these reasons, the classification and positional assignment we have given to NOD Vκ genes along the Igκ locus in this report is tentative, as the high degree of polymorphism in many sequences prevents fully confident assignment. We compile here the results of multiple different sequencing efforts in our lab centered around the NOD mouse, therefore the relative abundance of given light chains is likely due to sampling bias for light chains of particular binding specificities not discussed here, rather than to bias of amplification of particular Vκ genes.

As shown in Table 1, NOD Vκ genes which match with published NZBxNZW F1 sequences typically do not also match with MRL sequences, and vice versa. Furthermore, if the proposed IMGT-based gene identity is considered and NOD Vκ genes are arranged based on their proposed location in the Igκ locus, NZBxNZW F1 sequences which match NOD sequences localize to the 5′ portion of the locus, whereas MRL sequence matches localize to the 3′ portion of the locus, proximal to the Jκ segments (Fig. 5). Different strains of mice have been previously reported to possess different Igκ haplotypes (Kofler et al. 1989). It is tempting to speculate that the acquisition of these polymorphic regions may have occurred differently among the three strains. Sequencing of the Igκ locus has not been completed for NOD, NZBxNZW F1 or MRL strains of mice, thus the apparent positional bias could also be due to a bias in the sequences that have been deposited into GenBank. It is also possible that polymorphisms are not restricted to V genes in the NOD Igκ locus, but may also extend to regulatory elements. If so, this might influence the efficiency of processes such as receptor editing which are critical for tolerance maintenance. Polymorphisms in the recombination signal sequence have previously been recognized to affect recombination efficiency of particular Vκ genes, which ultimately influence host immunity (Atkinson et al. 1996; Feeney et al. 1996). Furthermore, MRL/lpr mice, which appear to share at least a portion of polymorphic Vκ genes present in the NOD locus, show reduced receptor editing (Lamoureux et al. 2007). Complete sequencing of the NOD Igκ locus will therefore be useful in identifying whether such additional differences are present.

We find polymorphisms in 43 independent Vκ genes spread throughout the Igκ locus, many of which are shared among SLE-prone strains of mice. The disease relatedness of these findings is questioned by the observation that neither the Igκ or TCR loci are present among the identified T1D disease susceptibility loci. However both B cell and T cell antigen specificity have been shown to play a key role in the development of T1D (Hulbert et al. 2001; Lennon et al. 2009; Silveira et al. 2002). It has previously been suggested that TCRα polymorphisms in the NOD mouse promote the formation of anti-insulin T cells to drive T1D (Simone et al. 1997b), and it is possible that the Vκ polymorphisms identified may similarly skew the B cell repertoire in the NOD strain. Multiple tolerance defects in the NOD mouse may fail to cull B cell autoimmunity, which combine with multigenetic influences throughout other cellular compartments to promote T1D. We speculate that the presence of an autoimmune Igκ haplotype might serve as a compounding risk factor for T1D development, and may thus potentially serve as a useful biomarker in identifying individuals who are at elevated risk for developing disease. However, further work is required to prove this possibility, as it is equally possible that the observed polymorphic Vκ genes may not directly contribute to enhanced disease development in NOD mice, but may have simply been inherited alongside IDD loci genes involved in disease progression.

Acknowledgments

This work was supported by National Institutes of Health (NIH) grants 5 T32 HL069765, 5 T32 GM08554, AI051448, K08 DK070924, and the Juvenile Diabetes Research Foundation (JDRF) 1-2008-108. We would like to acknowledge Guowu Yu, Mariam Karamali, and Allison Sullivan, as well as Vanderbilt Medical Center Flow Cytometry Shared Resource members Kevin P. Weller, Brittany Matlock, and David K. Flaherty [supported by the Vanderbilt Ingram Cancer Center (NIH P30 CA68485) and the Vanderbilt Digestive Disease Research Center (NIH DK058404)], the Vanderbilt DNA Sequencing Facility [supported by NIH CA68485, DK20593, and HL65962] for technical support.

Footnotes

Electronic supplementary material The online version of this article (doi:10.1007/s00251-010-0457-9) contains supplementary material, which is available to authorized users.

Contributor Information

Rachel A. Henry, Department of Medicine, Division of Rheumatology, Vanderbilt University, Nashville, TN, USA.

Peggy L. Kendall, Department of Medicine, Division of Allergy, Pulmonary, and Critical Care, Vanderbilt University, Nashville, TN, USA.

Emily J. Woodward, Department of Medicine, Division of Rheumatology, Vanderbilt University, Nashville, TN, USA.

Chrys Hulbert, Department of Medicine, Division of Rheumatology, Vanderbilt University, Nashville, TN, USA.

James W. Thomas, Department of Medicine, Division of Rheumatology, Vanderbilt University, Nashville, TN, USA.

References

  • Atkinson MJ, Cowan MJ, Feeney AJ. New alleles of IGKV genes A2 and A18 suggest significant human IGKV locus polymorphism. Immunogenetics. 1996;44:115–120. [PubMed]
  • Carrillo J, Puertas MC, Planas R, Pastor X, Alba A, Stratmann T, Pujol-Borrell R, Ampudia RM, Vives-Pi M, Verdaguer J. Anti-peripherin B lymphocytes are positively selected during diabetogenesis. Mol Immunol. 2008;45:3152–3162. [PubMed]
  • Feeney AJ, Atkinson MJ, Cowan MJ, Escuro G, Lugo G. A defective Vkappa A2 allele in Navajos which may play a role in increased susceptibility to haemophilus influenzae type b disease. J Clin Invest. 1996;97:2277–2282. [PMC free article] [PubMed]
  • Hulbert C, Riseili B, Rojas M, Thomas JW. B cell specificity contributes to the outcome of diabetes in nonobese diabetic mice. J Immunol. 2001;167:5535–5538. [PubMed]
  • Janeway CA, Jr, Travers P, Walport M, Shlomchik M. Immunobiology: the immune system in health and disease. New York, NY: Garland; 2001.
  • Kendall PL, Yu G, Woodward EJ, Thomas JW. Tertiary lymphoid structures in the pancreas promote selection of B lymphocytes in autoimmune diabetes. J Immunol. 2007;178:5643–5651. [PubMed]
  • Kofler R, Duchosal MA, Dixon FJ. Complexity, polymorphism, and connectivity of mouse Vk gene families. Immunogenetics. 1989;29:65–74. [PubMed]
  • Lamoureux JL, Watson LC, Cherrier M, Skog P, Nemazee D, Feeney AJ. Reduced receptor editing in lupus-prone MRL/lpr mice. J Exp Med. 2007 [PMC free article] [PubMed]
  • Lennon GP, Bettini M, Burton AR, Vincent E, Arnold PY, Santamaria P, Vignali DA. T cell islet accumulation in type 1 diabetes is a tightly regulated, cell-autonomous event. Immunity. 2009;31:643–653. [PMC free article] [PubMed]
  • Rojas M, Hulbert C, Thomas JW. Anergy and not clonal ignorance determines the fate of B cells that recognize a physiological autoantigen. J Immunol. 2001;166:3194–3200. [PubMed]
  • Silveira PA, Johnson E, Chapman HD, Bui T, Tisch RM, Serreze DV. The preferential ability of B lymphocytes to act as diabetogenic APC in NOD mice depends on expression of self-antigen-specific immunoglobulin receptors. Eur J Immunol. 2002;32:3657–3666. [PubMed]
  • Simone E, Daniel D, Schloot N, Gottlieb P, Babu S, Kawasaki E, Wegmann D, Eisenbarth GS. T cell receptor restriction of diabetogenic autoimmune NOD T cells. Proc Natl Acad Sci USA. 1997a;94:2518–2521. [PubMed]
  • Simone EA, Yu L, Wegmann DR, Eisenbarth GS. T cell receptor gene polymorphisms associated with anti-insulin, autoimmune T cells in diabetes-prone NOD mice. J Autoimmun. 1997b;10:317–321. [PubMed]
  • The Autoimmune Diseases Coordinating Committee, National Institutes of Health. Progress in Autoimmune Diseases Research. U.S. Department of Health and Human Services, National Institutes of Health National Institute of Allergy and Infectious Diseases; 2005. Ref Type: Report.
  • Thomas JW, Hulbert C. Somatically mutated B cell pool provides precursors for insulin antibodies. J Immunol. 1996;157:763–771. [PubMed]
  • Thomas JW, Kralick PM, Ewulonu UK. T cell-independent response to Brucella-insulin identifies a preimmune repertoire for insulin. J Immunol. 1997;159:2334–2341. [PubMed]
  • Thomas JW, Kendall PL, Mitchell HG. The natural autoantibody repertoire of nonobese diabetic mice is highly active. J Immunol. 2002;169:6617–6624. [PubMed]
  • Woodward EJ, Thomas JW. Multiple germline kappa light chains generate anti-insulin B cells in nonobese diabetic mice. J Immunol. 2005;175:1073–1079. [PubMed]
  • Yu L, Robles DT, Abiru N, Kaur P, Rewers M, Kelemen K, Eisenbarth GS. Early expression of antiinsulin autoantibodies of humans and the NOD mouse: evidence for early determination of subsequent diabetes. Proc Natl Acad Sci USA. 2000;97:1701–1706. [PubMed]