B cell proliferation, somatic hypermutation, class switch recombination, and autoantibody production in ectopic lymphoid tissue in murine lupus

Abstract

Intraperitoneal exposure of nonautoimmune mice to 2,6,10,14-tetramethylpentadecane (TMPD) causes lupus and the formation of ectopic lymphoid tissue. Although associated with humoral autoimmunity, it is not known whether Ab responses develop within ectopic lymphoid tissue or if B cells only secondarily migrate there. We show that ectopic lymphoid tissue induced by TMPD not only resembles secondary lymphoid tissue morphologically, but it also displays characteristics of germinal center reactions. Proliferating T and B lymphocytes were found within ectopic lymphoid tissue, activation-induced cytidine deaminase was expressed, and class-switched B cells were present. The presence of circular DNA intermediates, a hallmark of active class switch recombination, suggested that class switching occurs within the ectopic lymphoid tissue. Individual collections of ectopic lymphoid tissue ("lipogranulomas") from the same mouse contained different B cell repertoires, consistent with local germinal center-like reactions. Class-switched anti-RNP autoantibody-producing cells were also found in the lipogranulomas. Somatic hypermutation in the lipogranulomas was T cell-dependent, as was the production of isotype-switched anti-Sm/RNP autoantibodies. Thus, ectopic lymphoid tissue induced by TMPD recapitulates many of the functional characteristics of secondary lymphoid tissue and contains autoantibody-secreting cells, which may escape from normal censoring mechanisms in this location.

Figure 1. B and T cell proliferation in lipogranulomas

A, Immunohistochemistry of a TMPD-induced lipogranuloma (serial sections) demonstrating the presence of contiguous B cell (B220⁺) and T cell (CD3⁺) zones as well as cellular proliferation, as demonstrated by Ki-67 staining. Bottom right panel shows the absence of cells staining with the follicular dendritic cell marker FDC-M1 (FDC). B, Flow cytometry of lipogranuloma cells. Gates were set on either the B cells anti-CD45R (B220) or T cells (anti-CD4) and the percentage of cells staining with anti-Ki67 antibodies was determined. C, In vivo BrdU labeling of T and B cells in the lipogranulomas and spleens of TMPD-treated mice. Single cell suspensions were stained with anti-CD45R (B220), anti-CD3, and anti-BrdU antibodies. Data are expressed as the % of BrdU⁺ B cells or T cells, respectively.

Figure 2. TMPD lipogranulomas contain class switched B cells

A, Lipogranulomas express AID. Left, cDNA from TMPD or mineral oil lipogranulomas, spleen, or peritoneal cells was amplified using primers specific for AID or β-actin and analyzed by agarose gel electrophoresis. Right, AID mRNA was quantified by real-time PCR normalized to 18S RNA. B, μ and γ H-chain transcripts. Lipogranuloma and spleen cDNA from mineral oil or TMPD treated mice was tested for expression of IgM and IgG1 by PCR using VHF1-8 and Cμ or Cγ1 reverse primers, respectively. Left, agarose gel of amplified PCR products from lipogranulomas or spleen. γ1 transcripts are seen strongly in two TMPD lipogranulomas (TMPD, lanes 1 and 4) and weakly in another (TMPD, lane 3). Right, frequencies of IgM and IgG1 production in TMPD vs. mineral oil lipogranulomas (4 individual lipogranulomas/mouse, 3 mice/group). C, Circle transcripts. Presence of isotype specific I promoter-Cγ2a transcripts in lipogranulomas from TMPD-treated but not in mineral oil treated mice. Circle transcripts were detected using Iγ2aF forward primer and CμR reverse primer. PCR product sizes closely approximated the expected sizes of 458 and 318 bp (14) (arrows). PCR using B-actin primers was used as a loading control. D, Direct immunofluorescence for IgM (top) and IgG (bottom) producing cells in lipogranulomas from mineral oil or TMPD-treated mice (green). Nuclei were visualized by DAPI staining (blue). Both IgM and IgG producing cells were detected in ectopic lymphoid tissue from TMPD-treated mice, but only IgM producing cells in tissue from mineral oil-treated mice.

Figure 3. Individual TMPD-induced lipogranulomas contain distinctive populations of B cells

A, pooled forward primers (VHF1-8) and consensus reverse primer (VHR2) used for amplifying cDNA from lipogranulomas. B, H-chains recovered from lipogranulomas obtained from TMPD-treated BALB/c mice. Lipogranulomas #137, 139, and 140 were isolated from the same mouse. Lipogranulomas #190 and 193 were from a second mouse and lipogranulomas #201 and 204 from a third mouse. C, H-chains recovered from two lipogranulomas (#149 and 150) obtained from a mineral oil-treated BALB/c mouse.

Figure 4. V_H sequences from TMPD-induced lipogranulomas

A, Sequence alignments of VH36-60-DFL16.1-JH4 H-chains isolated from lipogranulomas #137 and 139 (two individual lipogranulomas from a single TMPD-treated mouse). Sequences obtained from the two different granulomas were unrelated, whereas the 6 sequences in granuloma #137 were identical. B, Sequence alignments of J558.f-DSP2.9-JH2 H-chains isolated from two different lipogranulomas (#149 and 150) from a mineral oil-treated mouse.

Figure 5. IgG1 and IgG2a induced hypergammaglobulinemia in TMPD-treated mice is T cell dependent

A, Serum samples were obtained from wild type C57BL/6J (WT, n = 5) or B6.129P2-Tcrb^tm1MomTcrd^Tm1Mom (n = 6) mice treated 3 months earlier with TMPD. IgG1, IgG2a, IgG3, and IgM levels were measured by ELISA and means were compared by the Mann-Whitney test. B, IgG production in lipogranulomas. Lipogranuloma cells and splenocytes from two mice were tested in quadruplicate for T cell dependent immunoglobulin secretion (IgG1+IgG2a+IgG2b) by ELISPOT assay. C, Quantification of plasmablasts in spleen and lipogranulomas. Pooled lipogranuloma and spleen cells from four TMPD-treated BALBc/J mice were stained with anti-B220 and anti-CD138 antibodies and analyzed by flow cytometry.

Figure 6. IgG anti-nRNP/Sm autoantibody production in TMPD-treated mice is T cell dependent

A and B, Serum samples were obtained from wild type C57BL/6J (WT, n = 5) or B6.129P2-Tcrb^tm1MomTcrd^Tm1Mom (n = 6) mice treated 3 months earlier with TMPD. IgM (A) and IgG (B) anti-nRNP/Sm antibody levels were measured by ELISA at a 1:500 serum dilution. Means were compared by the Mann-Whitney test. C, Reactivity of sera with recombinant U1-A protein (ELISA). Recombinant 6His-tagged U1-A protein was expressed in E. coli and purified on a Ni-NTA affinity column. Sera from 20 TMPD-treated BALB/c mice positive for anti-Sm/RNP autoantibodies and 20 normal BALB/c mouse sera were tested for reactivity with the recombinant antigen at a 1:100 dilution (ELISA). D, IgM and IgG ELISPOT assay with purified U1-A antigen using cells isolated from collagenase treated ectopic lymphoid tissue from an anti-RNP positive TMPD-treated mouse or an anti-RNP negative mouse treated with medicinal mineral oil. Representative of 3 experiments. E, IgG ELISPOT assay with purified U1-A or bovine serum albumin (BSA) antigens, using cells isolated from lipogranulomas (Lipogran) or spleens of anti-U1-A positive mice (n = 5). The frequencies of antigen-specific spots are expressed per 50,000 B cells. The frequency of anti-U1-A spots was higher in lipogranulomas than in spleen (P = 0.01, Mann-Whitney test) and the frequency of anti-U1-A spots was higher than the frequency of anti-BSA spots (P = 0.03, Mann-Whitney test).