Human CD57⁺ germinal center-T cells are the major helpers for GC-B cells and induce class switch recombination

We examined the distribution of T helper cell subsets in human tonsils based upon the expression of CD4, CD57 and CD69. As reported previously [12, 19, 28, 29], most CD57⁺ CD4⁺ T cells are located in germinal centers surrounded by IgD⁺ naïve B cells (Fig. 1). Small numbers of CD57⁺ T cells were also present in the interfollicular areas (IFA or T cell-rich zone) surrounding GC. Although some are found in IFA, CD69⁺ CD4⁺ T cells were also preferentially found in GC (Figure 1C). In contrast, the T cells in IFA were mostly negative for CD69 expression. Therefore, CD57, CD69 and CD4 are useful markers to identify CD57⁺ GC-Th cells and other T cell subsets differentially localized in tonsils: CD4⁺CD57⁺ cells (mainly in GC), CD4⁺CD57^-CD69⁺ cells (mainly in GC and a minor proportion in IFA), and CD4⁺CD57^-CD69^- cells (mainly in IFA).

Based upon the information obtained in Figure 1, the total tonsil CD4⁺ T cell population was fractionated into CD57⁺ GC-Th cells (all of these cells are CD69⁺), total CD57^- T cells, CD57^-CD69⁺ T cells and CD57^-CD69^- T cells (Figure 2). We compared the B cell helping activity of CD57⁺ GC-Th cells with that of other CD4⁺ T cell subsets. We co-cultured each of the isolated T cell subsets with syngeneic tonsil CD19⁺ B cells in the presence of SEB, a superantigen that conjugates MHC class II molecules and TCR (Figure 3). B cell receptors were cross-linked by Ab to human Ig μ chain and human Ig (H + L) chain prior to culture to provide BCR activation signals mimicking the antigen signals in vivo. CD57⁺ GC-Th cells were most efficient in inducing B cell production of IgM, IgG, IgA and IgE among the T cell subsets examined. CD57^- CD69⁺ T cells, many of which are located in GC in a manner similar to CD57⁺ GC-Th cells, were able to induce the production of antibodies but at significantly lower levels compared to CD57⁺ GC-Th cells (Figure 3A). T cell stimulation, in this study by SEB, was required for efficient induction of the B cell helper activity as it enhanced Ig production up to ~1000 (not shown).

Because of their specific localization in germinal centers, the physiological target cells for CD57⁺ GC-Th cells would be GC-B cells rather than naïve B cells. We compared the helper activities of CD57⁺ GC-Th cells and CD57^- CD69^+/- CD4⁺ T cell subsets for B cells. In this study, we fractionated CD19⁺ B cells into two groups: IgD⁺CD38^- naïve B and CD38⁺ IgD^+/- GC B cells as shown in Figure 2B. CD57⁺ GC-Th cells, when co-cultured with GC-B cells, were significantly more efficient than CD57^-CD69⁺ T cells in inducing the production of all four isotypes of Ig (Figure 3C). However, when co-cultured with naive B cells, CD57⁺ GC-Th cells were not significantly different from CD57^-CD69⁺ T cells in their induction capacity of Ig (Figure 3B). Again, the helper activities of total CD57^- T cells and CD57^-CD69^- T cells for naïve and GC-B cells were very low.

The relative composition of IgM, IgG, IgA and IgE produced in response to CD57⁺ GC-Th cells in the cultures with GC-B vs. naïve B cells was determined. CD57⁺ GC-Th cells drove the production of IgM, IgG, IgA and IgE in descending order (Figure 3D). Class-switched Ig isotypes such as IgG and IgA were more produced in GC-B cell cultures than in naïve B cell cultures. There was no statistically significant difference between the two T cell subsets (CD57⁺ GC-Th cells and CD4⁺CD57^-CD69⁺ T cells) in the composition of Ig that they induced.

AID expression in the maturating B cells in GC is necessary for CSR and somatic hypermutation. We examined whether CD57⁺ GC-Th cells have the capacity to induce AID in B cells. Naïve B cells were co-cultured with CD57⁺ GC-Th cells, and AID expression was examined (Figure 4A). CD57⁺ GC-Th cells induced AID in activated B cells with peaks around days 3–4. CD57⁺ GC-Th cells were able to induce the expression of productive V_HDJ_H-C_H Ig transcripts. The major subtypes of Ig transcripts in response to CD57⁺ GC-Th cells were IgG3, IgG2, IgG1 and IgA1 (Figure 4B). When the peak expression levels of AID and the productive V_HDJ_H-C_H IgG3 transcript (the most readily detected Ig transcript) were compared, AID expression preceded the expression of IgG3 transcript by 1–2 days in culture (Figure 4C).

Ig class switch recombination between tandemly repeated S regions located 5' to each C_H gene generates switch circles. We used a nested PCR technique designed to specifically detect switch circles but not genomic Ig sequences. Freshly isolated GC-B, but not naïve B cells, contained switch circles, which were detected as smeared multiple bands on agarose gels as expected. Naïve B cells cultured with CD57⁺ GC-Th cells generated detectable switch circles in a time-dependent manner (Figure 4D). We also used a DC-PCR technique [30] to detect γ3 and α1/2 switch circles (Figure 4E). Again, GC-Th cells induced switch circles in the naïve B cells cultured with GC-Th cells.

Cytokines and CD40L regulate B cell maturation and Ig production. We examined whether CD40L, IL-4, IL-10 and IFN-γ play any roles in the CD57⁺ GC-Th cell-driven B cell production of Ig. In cultures with naïve B cells, the blockage of CD40L by neutralizing antibody completely suppressed the helper activity of CD57⁺ GC-Th cells in inducing the B cell production of IgM, IgG1, IgA and IgE (Figure 5A). In this case, IgG1 was measured instead of total IgG to avoid cross-reaction of the polyclonal capturing antibody for IgG with the neutralizing antibodies to cytokines. Blockage of IL-4 partially but specifically suppressed the production of IgE, but it did not significantly suppressed other isotypes. In contrast, blockage of IFN-γ enhanced the production of IgM, IgG1 and IgA but not IgE. Since CD40L is essential for the helper activity of CD57⁺ GC-Th cells, we examined CD57⁺ GC-Th cells and other T cells for the expression of surface CD40L. Freshly isolated CD57⁺ GC-Th cells expressed CD40L, which became up-regulated upon T cell activation within hours (data not shown), whereas CD4⁺CD57^-CD69^- interfollicular T cells did not express CD40L at significant levels. There was no significant difference in the expression of surface CD40L between CD57⁺ and CD57^- CD69⁺ cells.

In the cultures with GC-B cells, blocking of CD40L again completely suppressed the B cell helping activity of CD57⁺ GC-Th cells (Figure 5B). However, IL-4 neutralization did not significantly affect the IgE production induced by CD57⁺ GC-Th cells, an activity different from that for naïve B cells. For GC-B cells, IFN-γ neutralization significantly increased the production of IgA as it did for naive B cells. The effects of IFN-γ neutralization on other Ig isotypes were smaller. While a slight decrease of IgE production in the cultures of GC-B cells and CD57⁺ GC-Th cells was observed, neutralization of endogenous IL-10 did not have any statistically significant effect on CD57⁺ GC-Th cell-driven Ig production in the cultures of either naïve or GC-B cells.

To further examine the regulatory effect of cytokines, IL-4, IL-10, IFN-γ and TGF-β1 were exogenously added to the cultures of CD57⁺ GC-Th cells with B cells (Figure 5C and 5D). In cultures of CD57⁺ GC-Th cells with naïve B cells, exogenously added IL-4 enhanced the production of some subsets of Ig, but this effect was small and not statistically significant (Figure 5C). However exogenously added IFN-γ significantly suppressed the production of IgG, IgA and IgE. IL-10, when added exogenously, was highly efficient in enhancing the production of the four subsets of Ig. TGF-β1 completely suppressed the B cell-helping capacity of CD57⁺ GC-Th cells for naive B cells.

In cultures of CD57⁺ GC-Th cells with GC-B cells, IL-10 was again highly effective in enhancing the helper activity of CD57⁺ GC-Th cells, while TGF-β1 completely suppressed it (Figure 5D). IFN-γ partially but significantly suppressed the production of IgM, IgG, IgA and IgE. Exogenous IL-4 added to the cultures had no effect on the CD57⁺ GC-Th cell-driven Ig production in this condition (Figure 5D), which is in line with the negligible effect of anti-IL-4 on GC-B cells in Figure 5B.

Mononuclear cells were prepared by density gradient centrifuge on histopaque 1077 (Sigma-Aldrich, St. Louis) from human tonsil pathological specimens obtained from young patients (3–10 yr) undergoing tonsillectomy to relieve obstruction of respiratory passages and improve drainage of the middle ear at Sagamore Surgical Center (Lafayette, IN). The use of human pathological specimens in this study was approved by the institutional review board at Purdue University. CD4⁺ T cells (purity >97%) were isolated by depleting non-CD4⁺ T cells using a magnetic bead depletion method (Miltenyi Biotec, Auburn, CA). After staining of the isolated CD4⁺ T cells with appropriate antibodies, CD57⁺ GC-Th cells (purity >95%) were isolated by a positive magnetic selection method (Miltenyi Biotec). CD4⁺CD57^-CD69⁺ and CD4⁺CD57^-CD69^- T cell subsets (purity >95%) were further isolated from the CD57^- T cell fraction by magnetically selecting CD69⁺ T cells (Miltenyi Biotec). Total B cells were isolated by rosetting with 2-amino-ethylisothiouronium bromide (AET)-treated sheep red blood cells followed by CD4⁺ T cell depletion (CD19⁺ cells > 99.5%). Naïve B cells (CD19⁺IgD⁺ cells >99%) were isolated from the total B cell fraction by depleting CD38⁺ T cells followed by positive magnetic selection of IgD⁺ B cells. CD19⁺CD38⁺IgD^+/- GC-B cells (purity >95%) were isolated from the tonsil CD19⁺ B cells as described before [57] using anti-CD44, anti-IgD antibodies and pan-mouse IgG beads (Dynal, Brown Deer, WI).

All cell cultures were performed in RPMI1640 medium supplemented with 10% FBS, gentamycin, streptomycin, and penicillin. To cross-link the B cell receptors, isolated B cells were incubated for 2 h at 4°C with Sepharose-conjugated rabbit Ab to human Ig μ chain and human Ig (H + L) chain (Irvine Scientific, Santa Ana, CA; mixed 1:1 at 2 μg/ml), and then washed with cold PBS. 10⁵ T and 10⁵ B cells were co-cultured, unless indicated otherwise, in each well of 48-well plates in the presence of Staphylococcal enterotoxin B (SEB; 1 μg/ml, Sigma-Aldrich, St. Louis, MO). Cells were incubated in 5% CO² incubators at 37°C for 3–8 days. Recombinant IL-4, IL-10, and TGF-β1 were purchased from R&D systems (Minneapolis, MN). Recombinant IFN-γ was obtained from BD Pharmingen (San Diego, CA). Purified CD154-blocking antibody (24–31) was obtained from Ancell Corporation (Bayport, MN). IL-4-blocking antibody (MP4-25D2) was purchased from BD Pharmingen. Blocking antibodies for IFN-γ (25718.111) and IL-10 (23738.111), and IgG1 isotype control antibody (11711.11) were purchased from R&D systems. All antibodies and reagents added to culture were azide-free. Cytokines were added at saturating concentrations: IL-4 (40 ng/ml), IL-10 (40 ng/ml), IFN-γ (200 ng/ml) and TGF-β1 (10 ng/ml). Neutralizing antibodies were added at following concentrations: anti-CD40L (20 μg/ml), anti-IL-4 (5 μg/ml), anti-IL-10 (5 μg/ml), anti-IFN-γ (2.5 μg/ml) and isotype antibody (5 μg/ml).

T cells were stained with anti-human CD57 (NK-1; FITC, BD Pharmingen), anti-human CD69 (FN-50; FITC, BD Pharmingen), anti-human CD4 (S3.5; R-PE, Caltag Laboratories, Burlingame, CA), and anti-human CD3 (UCHT1; APC, BioLegend, San Diego, CA). B cells were stained with anti-CD19 (4G7; PerCP, BD Pharmingen), anti-human IgD (IAb-2, FITC, BD Pharmingen), anti-human CD38 (HTT2; R-PE, BD Pharmingen), and anti-human CD3 (UCHT1; APC, BioLegend). Stained cells were analyzed using a 4-color FACSCalibur™ (BD Biosciences).

Frozen sections of tonsils were acetone-fixed and stained using antibodies to CD57 (BD Biosciences – Pharmingen; clone NK-1, labeled with FITC), CD69 (BD Biosciences – Pharmingen; clone FN50, labeled with FITC), IgD (BD Biosciences – Pharmingen; clone IA6-2, labeled with PE) and/or CD4 (Caltag Laboratories; clone S3.5, labeled with APC). Stained sections were analyzed using a confocal microscopy system (Bio-Rad MRC 1024UV and Nikon Diaphot 300 microscope) at Purdue Cytometry Lab.

Culture supernatants were assayed by ELISA as previously described [19]. The sensitivity of this ELISA system is greater than 5 ng/ml, 300 pg/ml, 30 pg/ml, 600 pg/ml, and 15 pg/ml for IgM, IgG, IgG1, IgA and IgE, respectively.

Total RNA was extracted from cultured cells with Trizol reagent (Invitrogen, Carlsbad, CA), and was reverse-transcribed into cDNAs with SuperScript™ First-Strand Synthesis System for RT-PCR (Invitrogen) according to the manufacturer's protocol. The primer pairs used in this study were designed by Cerutti et al. [37]: IgM, FR3 forward (5'-GAC ACG GCT GTG TAT TAC TGT GCG-3') and Cμ reverse (5'-CCG AAT TCA GAC GAG GGG GAA AAG GGT T-3'); IgG1, FR3 forward and Cγ1 reverse (5'-GTT TTG TCA CAA GAT TTG GGC TC-3'); IgG2, FR3 forward and Cγ2 reverse (5'-GTG GGC ACT CGA CAC AAC ATT TGC G-3'); IgG3, FR3 forward and Cγ3 reverse (5'-TTG TGT CAC CAA GTG GGG TTT TGA GC-3'); IgG4, FR3 forward and Cγ4 reverse (5'-ATG GGC ATG GGG GAC CAT TTG GA-3'); IgA1, FR3 forward and Cα1 reverse (5'-GGG TGG CGG TTA GCG GGG TCT TGG-3'); IgA2, FR3 forward and Cα2 reverse (5'-TGT TGG CGG TTA GTG GGG TCT TGC A-3'); IgE, FR3 forward and Cε reverse (5'-CGG AGG TGG CAT TGG AGG-3'); human β-actin, actin forward (5'-ATG TTT GAG ACC TTC AAC AC-3') and actin reverse (5'-CAC GTC ACA CTT CAT GAT GG-3'). PCR reactions were performed on serially diluted cDNA samples using an Eppendorf master cycler (denaturation at 95°C for 15 s, annealing at 55°C for 45 s and extension at 72°C for 30°C; 30–35 cycles). Extrachromosomal switch circles were detected by a nested PCR strategy as previously described by others [37]. Briefly, genomic/extrachromosomal DNA was isolated from fresh or cultured B cells using a QIAamp DNA Mini Kit (Qiagen, Valencia, CA) and was used as templates for amplification of Sγ1-Sμ, Sγ2-Sμ, Sγ3-Sμ, Sγ4-Sμ and Sα-Sγ. The PCR products were subject to second PCR using internal forward 5' Sγ or 5'Iα1/2i and reverse 3'Sμi or 3'γi primer pairs. This method has been verified for specificity using positive controls [37]. Additionally, we amplified genomic β-actin gene as a control using 5'-GTA CCA CTG GCA TCG TGA TGG ACT-3' (G-actin-forward-1 primer) and 5'-ATC CAC ACG GAG TAC TTG CGC TCA-3' (G-actin-reverse-1) for the first PCR; and 5'-AGA AGA GCT ACG AGC TGC CTG AC-3' (G-actin-forward-2) and 5'-TGA GGA CCC TGG ATG TGA CAG CT-3' (G-actin-reverse-2) for the second PCR. Additionally, we used a DC-PCR technique [30, 58] to demonstrate the presence of switch circles (γ3 and α1/2) in human B cells. Please see the reference [30] for primer sequences.

Total RNA was extracted from freshly isolated or cultured cells using Trizol reagent (Invitrogen, Carlsbad, CA), and was reverse-transcribed into cDNAs with SuperScript™ II Reverse Transcriptase. RT-PCR amplification of AID was performed using the two primers: AID-forward (5'-GAT GAA CCG GAG GAA GTT TC-3') and AID-reverse (5'-TCA GCC TTG CGG TCC TCA CAG-3'), which generated a specific 351 bp PCR product after 30 cycles of PCR reaction (30 s at 94°C, 30 s at 60°C, and 60 s at 72°C). β-actin was also amplified as a control.

Student's paired t-test was used. P values smaller than 0.05 were considered significant.