HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fear, D. J. Articles by Gould, H. J. Search for Related Content PubMed PubMed Citation Articles by Fear, D. J. Articles by Gould, H. J.

Transcription of Ig Germline Genes in Single Human B Cells and the Role of Cytokines in Isotype Determination¹

David J. Fear^*,, Natalie McCloskey^*, Brian O’Connor, Gary Felsenfeld and Hannah J. Gould^2,*

^* The Randall Center, King’s College London, United Kingdom; Laboratory of Molecular Biology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD 20892; and Department of Respiratory Medicine and Allergy, Guy’s, King’s, and St. Thomas’ School of Medicine, London, United Kingdom

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
We have developed a critical test of the chromatin accessibility model of Ig isotype determination in which local unfolding of chromatin higher order structure (chromatin accessibility) in the region of specific germline genes in the H chain locus determines the Ab class to be expressed in the B cell. We show that multiple germline genes are constitutively transcribed in the majority of naive human B cells in a population. Thus, because chromatin in its higher order structure cannot be transcribed, the entire Ig H chain locus must be unfolded in naive B cells. We have also established that IL-4 and anti-CD40 act by enhancing transcription in the majority of cells, rather than by activating transcription in more of the cells. Transcriptional activity in the human H chain locus rules out the perturbation of chromatin higher order structure as a factor in isotype determination. We have also found that the levels of germline gene transcription cannot fully account for the levels of secretion of the different Ig isotypes, and that secretion of IgE, in particular, is suppressed relative to that of IgG.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The adaptive immune response is characterized by two somatic DNA recombination events. The first, V(D)J recombination, determines the Ag specificity of an Ab, whereas the second, class switch recombination (CSR), ³ is required to alter the Ab’s effector function (1). Class switching to IgE is an event that may lead to allergic hypersensitivity (2), but the mechanism by which IgE is selected for expression in a B cell is not well understood. It has therefore been the focus of many studies, with the underlying assumption that a better understanding may lead to improved treatments.

There are nine Ab classes, distinguished from each other by the constant regions of their H chains, which are encoded in a tandem array (5'-Cµ, C, C3, C1, C1, C2, C4, C, C2–3') on human chromosome 14 (1). The variable (VDJ) region of the Ig H chain is first linked to Cµ and C to express the µ-chain of IgM or the -chain of IgD by differential splicing of a common mRNA precursor. Ag binding to the membrane-bound Ab together with cytokine and T cell stimulation promote switching to another Ab class by recombination of the VDJ region with one or, sequentially, two or more of the downstream CH genes (3).

The process of H chain class switching proceeds in three distinct stages: germline gene transcription, DNA recombination (CSR), and B cell differentiation into Ig-secreting plasma cells or memory B cells (2, 4). Each CH germline gene comprises a short germline exon (I), a repetitive GC-rich switch (S) region, and several exons encoding the H chain constant region (CH). Germline gene transcription is initiated at the I exon promoter and proceeds through the S region and the CH exons. This product undergoes splicing, removing the S region, to produce the mature (sterile) germline gene transcript (GLT) (1). Germline gene transcription always precedes and is necessary for DNA recombination (3).

CSR involves DNA recombination between the S regions of two CH genes, the originally expressed CH gene, linked to the VDJ region, and one of the downstream CH genes. The intervening DNA is looped out and excised from the genome. CSR has been shown to require participation of the recently discovered enzyme, activation-induced cytidine deaminase (AID) (5), which is also involved in somatic hypermutation and gene conversion, suggesting that there is a step common to all three mechanisms (6). The presence of uracil residues, resulting from the cytosine deamination, leads to DNA chain cleavage by uridine nucleotide glycolase, followed by recombination through the nonhomologous end-joining pathway (7, 8).

There is now considerable evidence that GLT may play an active role in the mechanism of CSR (9). GLT may be retained in actively transcribing chromatin, in the form of RNA hybrids with the coding strand of the DNA, forming so-called R-loops (10). The extended single-stranded region of the noncoding strand may then form hairpin loops, favored by palindromic sequences in the S regions (11, 12), which may be targeted by AID and uridine nucleotide glycolase (6, 9); recombination may also require GLT splicing (13, 14). After CSR, the B cell must differentiate into a plasma cell before Ig secretion (15, 16). In principle, isotype determination in the B cell population may occur at any stage of the class-switching process.

Knowledge of the mechanism of recombination does not necessarily shed light on the mechanism of isotype determination. Early work on murine B cells revealed that IL-4 selectively stimulates and 1 germline gene transcription before class switching to IgE and IgG1, whereas class switching to IgG2b follows when IFN- stimulates 2b germline gene transcription and is accompanied by suppression of the alternative switching pathways (17, 18, 19, 20, 21, 22). These findings led to the long-held chromatin accessibility model for isotype determination (3, 23, 24, 25), in which local unfolding of chromatin higher order structure initiates specific transcription and recombination of a specific germline gene, leading to synthesis and secretion of the Ab. This originally plausible model has not been rigorously tested, however.

By contrast with the situation in murine B cells, germline gene transcription and class switching to all the Ig isotypes are induced by IL-4 and anti-CD40 in human B cells (26, 27, 28, 29, 30). This implies that either B cells are a mixture of discrete subpopulations of cells that are committed to express a particular isotype, or every cell expresses all isotypes. If the latter model is correct, then the entire Ig H chain locus on chromosome 14 would have to be unfolded in all B cells, because DNA that is packaged in a higher order chromatin structure cannot be transcribed. Transcription of multiple germline genes in the same cell is thus inconsistent with the chromatin accessibility model.

The accessibility model could be modified to accommodate the apparent lack of specificity of cytokine-induced germline gene transcription if cytokines, such as IL-4, were to modulate the proportion of cells expressing any particular isotype. Thus, IL-4 should increase only the proportion of cells expressing GLT (or 4 GLT) in human B cells. To examine this possibility and to clarify the mechanism of isotype determination, we have measured germline gene transcription in single human tonsil B cells incubated with IL-4 and other cytokines and compared the results to those obtained with B cell populations.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cell purification

Naive human B cells were isolated from tonsils of donors undergoing routine tonsillectomies (ethical approval from Guy’s, King’s, and St. Thomas’ Hospital Trust). The patients were uncharacterized with respect to allergies, and no patient history was taken. Cells were isolated from the tonsil by maceration and washing in medium, and mononuclear cells were separated by density on a Ficoll gradient. Naive B cells were isolated by magnetic separation as follows. Mononuclear cells were incubated with a PE-conjugated anti-human IgD Ab (Southern Biotechnology Associates, Birmingham, AL) for 20 min, and the cells were washed and incubated with magnetic microbeads conjugated to an anti-PE Ab (Miltenyi Biotec, Bergisch Gladbach, Germany). Finally, IgD⁺ cells were analyzed for B cell purity by flow cytometry using a FACSCalibur flow cytometer (BD Biosciences, Oxford, U.K.). Cells were analyzed for IgD expression by PE staining and for surface Ig expression using FITC-conjugated IgG and IgA (DakoCytomation, Ely, U.K.) or IgE (Vector Laboratories, Burlingame, CA) and the markers, FITC-conjugated CD27 and CD69 (DakoCytomation). Isotype controls were gated at 2%; therefore, values <2% for the stained cells are classed as negative. Cell populations were routinely >95% negative for IgG-, IgA-, or IgE-expressing cells; CD27⁺ memory B cells; and CD69⁺-activated B cells (Fig. 1, A and B).

View larger version (31K): [in this window] [in a new window]   FIGURE 1. Phenotypic analysis of purified naive B cell populations. Naive human B cells were isolated from tonsillar tissue by MACS by virtue of surface IgD expression using a PE-conjugated anti-human IgD Ab. The purity of the isolated cell population was subsequently assessed by flow cytometry and RT-PCR. A, Representative analysis of residual IgD/PE staining compared with surface IgG, IgE, and IgA staining immediately after cell isolation. B, Representative analysis of IgD/PE staining compared with CD27 and CD69 expression, markers of memory and activated B cells, respectively. For the analysis of all FACS data, isotype controls were gated at 2%; quadrant values below this figure are taken as negative. C, RT-PCR analysis of AID expression in naive cells immediately after isolation, after 6 days culture without exogenous cytokines (unstimulated), or after 6 days stimulation with IL-4/anti-CD40. For each condition, RT-PCR was performed on 2 µl of a serial dilution of cDNA (1/1, 1/10, and 1/100, as indicated by triangles).

IgE^– B cells were isolated using a depletion-based protocol (B cell isolation kit; Miltenyi Biotec). Briefly, mononuclear cells from tonsils were incubated with a mixture of haptenated Abs (CD2, IgE, CD4, CD11b, CD16, and CD36, supplied in the isolation kit). Non-B cells were then magnetically labeled with MACS microbeads coupled to an anti-hapten Ab. Unlabeled cells were isolated using depletion columns (Miltenyi Biotec).

K562 cells (31) were obtained from American Type Culture Collection (Manassas, VA) and were grown in suspension in DMEM supplemented with antibiotics and 10% FBS (HyClone, Perbio Biosciences, Tattenhall, U.K.) as described previously.

T cells were isolated from PBMC, isolated as described above, by magnetic bead separation. Mononuclear cells were incubated with a PE-conjugated anti-human CD3 Ab (DakoCytomation) for 20 min. The cells were washed and incubated with magnetic microbeads conjugated to an anti-PE Ab (Miltenyi Biotech). Finally, CD3⁺ cells were analyzed for T cell purity and B cell contamination by flow cytometry using a FACSCalibur flow cytometer (BD Biosciences, Oxford, U.K.). T cell populations were found to be >95% pure with <2% contaminating B cells (data not shown).

Cell culture

All cells were cultured at 0.5 x 10⁶ cells/ml in Yssel’s medium (32) supplemented with antibiotics and 10% FBS (HyClone, Perbio Biosciences). Where indicated, media were also supplemented with 1 µg/ml anti-CD40 Ab (G28.5; American Type Culture Collection), 200 U/ml IL-4 (R&D Systems, Abingdon, U.K.), 1 ng/ml IFN- (PeproTech, London, U.K.), or 100 ng/ml IL-10 (PeproTech). Unless specified in the text, cells were cultured for 24 h before RT-PCR analysis and for 10 days for Ig production analysis.

RNA extraction and RT-PCR

RNA was extracted from 2 x 10⁶ cells using a phenol/chloroform/guanidinium isothiocyanate-based extraction kit (RNAwiz; Ambion, Austin, TX). Five micrograms of total RNA was subsequently primed with oligo(dT) and reverse transcribed using Superscript RT II (Invitrogen Life Technologies, Paisley, U.K.). Quantitative RT-PCR was performed by amplifying a serial dilution of cDNA using the F2/R2 primer sets (shown below) at 59°C. Samples were separated on a 1.5% agarose/Tris-borate/EDTA gel, blotted onto nylon membranes, and hybridized to specific probes (all internal of primers) before quantification on a phosphor imager (Bio-Rad, Hercules, CA).

To ensure that the PCR analyses were conducted in the quantitative region of PCR amplification, i.e., when band intensity was proportional to the amount of target input, intensities were examined as a function of dilution. Fig. 2C shows that all PCRs, with the exception of , gave a good correlation between band intensity and dilution in the 1/1 to 1/10 dilution range; in most cases, a 1/100 dilution gave an immeasurably small intensity. It can be seen that over the 10-fold dilution range, band intensity decreased by a factor of 5. Thus, the GLT levels quoted represent a lower limit of 2-fold underestimation of actual values. In the case of , it can be seen that for the IL-4/anti-CD40-stimulated samples band intensity does not accurately reflect cDNA dilution; this demonstrates that the PCR had reached saturation within the dilution range. For the samples with a lower level of GLT expression (anti-CD40-stimulated samples), the results were not vitiated by saturation. Therefore, we stated lower limits of GLT stimulation by IL-4 and anti-CD40, based on the band intensity at the highest dilution.

View larger version (47K): [in this window] [in a new window]   FIGURE 2. Semiquantitative RT-PCR analysis of germline transcripts in naive human B cells cultured for 24 h. A, Naive human B cells were isolated from tonsillar tissue and cultured for 24 h in IL-4, IL-4 and IL-10, IL-10, and IFN- (all in the presence of anti-CD40), anti-CD40 alone, or without exogenous cytokine stimulation (unstimulated). For each condition, RT-PCR was performed on 2 µl of a serial dilution of cDNA (1/1, 1/10, and 1/100, as indicated by triangles) and subjected to 25 cycles of PCR amplification (20 cycles for GAPDH). Results were analyzed by phosphorimager analysis and corrected to a constant GAPDH signal. The fold stimulation of the level of each transcript is compared with that of the cells cultured without exogenous cytokine stimulation. B, Naive human B cells were isolated from peripheral blood taken from a normal healthy donor and cultured for 24 h in IL-4/anti-CD40, anti-CD40 alone, or without exogenous cytokine stimulation. RT-PCR was conducted as described in A. In some cases, at the highest concentration the PCR generated additional nonspecific bands. Only the intensities of the specific bands, labeled with an arrow, were analyzed for subsequent quantitation. C, To ensure that the RT-PCRs were analyzed in the linear phase of the PCR, we plotted band intensity against cDNA dilution. For each sample and primer set, band intensity is expressed as a fraction of the band intensity obtained from the undiluted cDNA. All results shown (with the exception of , see below) are derived from the naive tonsil B cells stimulated with IL-4/anti-CD40 (shown in A): , GAPDH; , 1; , 2; , 3; and •, 4. In the case of GLT, the data shown are from the anti-CD40-stimulated cells () and IL-4/anti-CD40-stimulated cells (). D, To ensure the RT-PCR methods detected GLT with similar efficiencies, known quantities of cloned GLT cDNA targets were diluted and subjected to PCR amplification.

The method of single cell RT-PCR was modified from that described by Hu et al. (33). Cells were isolated and cultured as described above. Single cells were obtained by limiting dilution in PBS (1 cell/1 µl); 1 µl of cell suspension was pipetted into wells in 96-well plates, and wells with single cells were identified by microscopy. Cells were lysed for 15 min on ice in 4 µl of lysis solution (0.4% Nonidet P-40, 60 µM dNTPs, 1 µM DTT, and 0.5 U/µl RNasin). RT solution (6 µl) was added (1x first-strand buffer, 50 ng oligo(dT), 40 µM dNTPs, 5 µM DTT, and 100 U Superscript RT II), and the mixture was incubated for 1 h at 42°C. After RT, 2 µl of cDNA was amplified by two nested PCRs using the F1/R1 primers at 68°C, followed by F2/R2 primers at 59°C. Samples were separated by electrophoresis and blotted as described above. Primers were designed to ensure that only mature GLT were amplified. Amplified products exhibited the expected size of fully spliced GLT and hybridized to a probe located between the primer sequences in Southern blots. The use of fully nested primer sets, spanning intron and S regions removed by splicing, and the use of internal probes prevented the amplification of genomic DNA sequences and ensured product specificity.

RT-PCR primers

The following primers were used: GAPDH: F1, GGG GAA GGT GAA GGT CGG AGT C; F2, ATT TGG TCG TAT TGG GCG CCT GGT C; R1, CCT CAC CTG ATG ATC TTG AGG CTG TTG TCAT; and R2, TCA TAC TTC TCA TTG TTC ACA CCC ATG; : F1, GGG AGC TGT CCA GGA ACC CGA CAG AGC; F2, GG CCA CAC ATC CAC AGG C; R1, CAG GAC GAC TGT AAG ATC TTC ACG; and R2, GG GGT GAA GTC CCT GGA GC; : F1, AAG CCA ACA GGG CAG CAC ACA C; F2, GAT GCC AGG ATG GGC ACG AC; R1, GGG TTT TGG GGG GAA GAG GAA GAC; 1R2, GGC ATG TGT GAG TTT TGT CAC AA; 2R2, GTG GGC ACT CGA CAC AAC; 3R2, CCA AGT GGG GTT TTG AGC; and 4R2, GGG CAT GGG GGA CCA TA; AID: F1, TAG ACC CTG GCC GCT GCT ACC; and R1, CAA AAG GAT GCG CCG AAG CTG TCT GGA G (based on AID primers published previously (34)); and globin: F1, GGA GAA GGC TGC CGT CAC TAG C; R1, GCC TGC ACT TCA GGG GTG AAC TC; F2,CAA GAT GAA TGT GGA AGA GG; and R2, AAT AAT CAC CAT CAC GTT ACC C.

Detection of secreted Igs

Secreted Igs were assayed by ELISA after 10 days of cell culture.

IgE. Maxisorp plates (Nalge Europe, Neerijse, Belgium) were coated with polyclonal mouse anti-human IgE (DakoCytomation) diluted 1/7000 in sodium carbonate buffer, pH 9.8, overnight at 4°C. Unbound sites were then blocked with 2% nonfat milk powder (Marvel; Premier International Foods, Spalding, U.K.) in PBS/0.05% Tween for 30–60 min at room temperature. Samples were added at appropriate dilutions, and the plates were incubated for 16 h at 4°C; NIP-IgE (JW8/5/13; European Collection of Cell Cultures, Porton Down, U.K.) was used to construct a standard curve. IgE binding was detected by mouse anti-human IgE conjugated to HRP (DakoCytomation) diluted 1/1000 in PBS/0.05% Tween 20/1% nonfat milk powder for 4 h at room temperature. The color reaction was developed with OPD (Sigma-Aldrich). Typically the ELISA was sensitive down to 4 ng/ml.

IgG subclasses. Maxisorp plates were coated with polyclonal goat anti-human IgG (Oxford Biotechnology, Oxon, U.K.) diluted 1/1000 in carbonate buffer, pH 9.8, overnight at 4°C. Free sites were then blocked with 2% nonfat milk powder in PBS/0.05% Tween for 30–60 min at room temperature. Samples were added at appropriate dilutions, and the plates were incubated for 16 h at 4°C; IgGs of known subclasses (Sigma-Aldrich) were used to construct standard curves. IgG binding was detected by sheep anti-human IgG subclass-specific Abs conjugated to HRP (The Binding Site, Birmingham, U.K.) diluted 1/1000 for anti-IgG1, anti-IgG2, and anti-IgG4 and 1/3000 for anti-IgG3 in PBS/0.05% Tween 20/1% nonfat milk powder for 4 h at room temperature. The color reaction was developed with OPD (Sigma-Aldrich). Typically the ELISA was sensitive down to 4 ng/ml.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Phenotypic analysis of isolated naive B cell populations

We used naive human tonsil and peripheral blood B cells isolated by positive selection for the expression of IgD. Such cells have not yet undergone class switching and are therefore suitable for examining the effects of cytokines and anti-CD40 on the patterns of germline gene transcription and switching to the different Ig classes. A typical FACS analysis (shown in Fig. 1) demonstrates that after positive selection, the B cell populations were routinely >98% IgD⁺. That these cells had not previously undergone isotype switching was confirmed by the presence of <2% contaminating IgA⁺, IgG⁺, and IgE⁺ cells (Fig. 1A). Because isotype control gates were set at 2%, this implies an effectively negative character.

We examined whether the naive IgD⁺ B cell populations used in our work might have contained a significant proportion of B cells activated in situ or memory B cells that had not yet lost their IgD expression or their fully expressed isotype-switched membrane Ig. The inclusion of a significant proportion of such cells would skew further analysis toward an activated phenotype. Therefore, positively selected IgD⁺ cells were also tested for the expression of CD69, a marker of activated B cells (35), and CD27, a marker of memory B cells (36). FACS analysis revealed that the positively selected IgD⁺ population routinely contained <5% CD69⁺ and <2% CD27⁺ B cells (Fig. 1B). To ensure that the cells had not been stimulated in situ before isolation, we analyzed the expression of AID, a standard marker for cells that have entered the early stages of CSR (34). No AID expression was detected in the cells immediately after isolation or after 6, 7, and 8 days in culture without exogenous cytokine stimulation. By contrast, AID was readily detected at these later time points in cells stimulated with IL-4 and anti-CD40 (Fig. 1C). Thus, the majority of cells in the isolated population were truly naive, as judged by all available criteria.

RT-PCR analysis of germline gene transcription in B cell populations

A semiquantitative RT-PCR analysis of GLT levels was conducted on the naive tonsil B cell population (Fig. 1A), which was cultured for 24 h in the presence of IL-4, IL-10, IFN-, or a combination of IL-4 and IL-10, all in the presence of anti-CD40. For comparison, cells were cultured with anti-CD40 alone or without exogenous cytokine stimulation. In all cases a serial dilution (1/1, 1/10, and 1/100) of cDNA was analyzed for the expression of , 1, 2, 3, and 4 GLT, and the results for each product were ratioed to the expression of the housekeeping gene GAPDH. Products from each primer set have been previously sequenced to ensure amplification of the correct target sequence (results not shown). Due to the high degree of homology between the genes, we were concerned that the reverse primers might cross-react. Cloned products from each of the gene PCRs were used as templates for the other gene PCRs; no significant cross-reactivity was observed (results not shown).

Absolute levels of GLT expression and stimulation were found to vary between donors (n = 4; results not shown), but Fig. 2A is representative of the observed patterns of expression. These results show that 1, 3, and GLT are constitutively expressed at a significant level in naive B cells. 2 and 4 GLT appear to be expressed at a lower level, as indicated by the higher cycle numbers and longer exposures required for their appearance (results not shown). Similar results were obtained with IgD⁺ B cells isolated from PBMCs (Fig. 2B), thus confirming that the results obtained with IgD⁺ tonsil B cells were not due to stimulation in their lymphoid environment.

In the example shown in Fig. 2A, 1 GLT were stimulated only weakly (3-fold), from a high basal level, by the addition of IL-4/anti-CD40. 3 GLT were significantly stimulated (20-fold) by IL-4/anti-CD40, IL-4 with IL-10/anti-CD40 (18-fold), and, to a lesser extent (3-fold), IL-10/anti-CD40 alone. 2 GLT were also stimulated by IL-4/anti-CD40 (8-fold), IL-4 with IL-10/anti-CD40 (6-fold), and, again to a lesser extent, IL-10/anti-CD40 alone (2.5-fold). 4 GLT were very barely detectable in the unstimulated cells cultured without exogenous cytokines, but were readily detected in the IL-4/anti-CD40-stimulated (>100-fold stimulation over medium alone) or IL-4- with IL-10/anti-CD40-stimulated cells (>50-fold stimulation over unstimulated cells). Under all conditions, some level of GLT expression was observed. However, IL-4/anti-CD40 and IL-4 with IL-10/anti-CD40 strongly increased transcription by >150- and 25-fold, respectively, compared with unstimulated cells. Despite variations in the levels of stimulation for each GLT with the different cytokines, in all cases IL-4/anti-CD40 or IL-4 plus IL-10/anti-CD40 gave the maximal levels of GLT stimulation compared with other conditions.

Although absolute comparisons of the level of expression of the different H chain genes cannot be made by this procedure, we ensured that all data were analyzed in the quantitative region of the PCR (see Materials and Methods; Fig. 2C). Furthermore, the efficiencies of all PCRs were approximately equal; all primer sets, with the exception of 4, could detect similar levels of their diluted targets and generated bands of comparable intensity as a function of template concentration (Fig. 2D). The detection sensitivity of the 4 primer set was 10 times less than that of all others, accounting for the weaker signals. Thus, we can approximate the levels of expression of the different GLT in B cells incubated under the same conditions. With the exception of 2, the maximal levels of expression of all GLT were very similar. The constitutive levels of , 1, and 3 GLT expression were also very similar, whereas GLT was expressed at a 10 times higher level than 1, 3, and 4 GLT upon stimulation with IL-4/anti-CD40.

Cytokine stimulation of Ig secretion in B cell populations

Table I shows the levels of IgE, IgG1, IgG2, IgG3, and IgG4 secreted from the naive tonsil B cell population (shown in Figs. 1 and 2) after 10 days in culture. Absolute levels of secreted Igs varied among the three donors examined (results not shown), but the relative amounts of the different Ig classes were similar to those shown in Table I. With the exception of IgG2, similar results were obtained with naive, IgD⁺ peripheral blood B cells (see below). As in previous work (27, 29), cytokine-stimulated Ig secretion was compared with secretion from cells cultured without exogenous cytokines. As usual, cell death was somewhat greater in the latter populations, but the majority of cells remained viable, as judged by trypan blue dye exclusion (data not shown).

These results (Fig. 2 and Table I) confirm previous findings that IL-4 is not specific for IgE and IgG4 secretion in human B cells (26, 27, 28, 29, 30) and crucially establish that IL-4 is at least as effective in promoting production of IgG isotypes as IgE.

Distribution of GLT in single B cells

We have shown that B cell populations express multiple GLT constitutively and that IL-4/anti-CD40 stimulates GLT expression. To relate these results to the behavior of individual cells in the population, we asked whether constitutive GLT expression is confined to a small proportion of the cells, and whether IL-4 selects different, precommitted B cell subpopulations or instructs subpopulations of the B cell population at random to express specific isotypes.

An analysis of GLT expression in single naive tonsil B cells is shown in Fig. 3A. Data are for a single set of 21 sorted cells plus negative controls. The indicated percentage of -positive cells represents results from a minimum of three donors (total n = 63). Surprisingly, these results show that a large proportion (41%) of unstimulated cells express GLT, and that this is unchanged, within experimental error, upon stimulation by IL-4, IL-10, or IFN- in the presence of anti-CD40. The SDs derived from the accumulated data show that the absolute numbers of cells expressing GLT varied from donor to donor. In most cases, the GLT signal was detected only in cells displaying a GAPDH signal, although occasionally only an signal was detected, probably due to a failure in the GAPDH PCR. These cells are omitted from the -positive count. Thus, 41% was the minimum of -positive cells detected by this procedure.

View larger version (33K): [in this window] [in a new window]   FIGURE 3. Analysis of GLT expression in single B cells. A, Analysis of GLT expression in single naive tonsillar B cells. Naive B cells were isolated from tonsils and cultured for 24 h in IL-4, IL-10, or IFN- (all in the presence of anti-CD40) or without exogenous cytokine stimulation (unstimulated). Single cells were isolated by limiting dilution on a 96-well plate, and wells containing single cells were identified microscopically before cell lysis and RT-PCR analysis. One representative set of results (21 cells from a single donor plus negative controls) is shown for each culture condition. Experiments were not analyzed unless all negative controls were clear. The percentages of -positive cells shown (±SD) were calculated from a minimum of three different donors (n = 63). Only cells showing a positive GAPDH signal were included in the numerical analysis of the percentage of -positive cells. B, Analysis of GLT expression in single B cells isolated by depletion. IgE– total B cells were isolated from tonsil tissue by a depletion, negative selection protocol. Single B cells were isolated, lysed, and subjected to RT-PCR analysis immediately after isolation. The percentage of -positive cells was calculated from two different donors (n = 42). C, Analysis of GLT in single naive peripheral blood B cells. Naive B cells were isolated from peripheral blood collected from a normal, healthy donor and cultured for 24 h in IL-4/anti-CD40 or without exogenous cytokine stimulation. Single B cells were then isolated, lysed, and subjected to RT-PCR analysis. We show the results from a single donor (n = 21). D, Analysis of GLT in single U266 cells. Cells of the U266 B cell line were subjected to single cell analysis. The results are representative of three experiments (n = 30).

To confirm the validity of the positive results, we additionally studied single cells of the U266 cell line. The U266 cell line has undergone CSR to express IgE, lacks the second allele of the germline gene (37), and is therefore unable to express GLT. GAPDH transcripts were present, but no GLT were detected in these cells (Fig. 3D), validating the significance of these transcripts in primary cells.

Semiquantitative analysis of germline gene transcription in single B cells

The single cell analysis shown in Fig. 3 demonstrates that a large proportion of cells actively transcribed the germline gene under all culture conditions, and that addition of IL-4 did not increase this proportion. However, the semiquantitative RT-PCR analysis indicated a strong enhancement of germline gene transcription in the naive B cell population as a whole (Fig. 2, A and B). Thus, it appears that the up-regulation of germline gene transcription by IL-4 was due to an increase in the number of transcripts per cell in the majority of cells rather than an increase in the number of expressing cells.

Due to the use of two rounds of nested PCR and the large number of cycles used to ensure maximum product detection, the results shown in Figs. 3 and 5 (see below) do not give quantitative information about the transcript levels in each cell. To confirm the inference of increased GLT expression after IL-4 and anti-CD40 stimulation, a semiquantitative single cell RT-PCR method was devised. During the first round of PCR the products were sampled at different stages of amplification (14, 18, 22, and 26 cycles). Each of these was then subjected to the second round of PCR, using nested primers for 18 cycles, and analyzed as before.

View larger version (75K): [in this window] [in a new window]   FIGURE 5. Analysis of and 1–4 GLT expression in single naive B cells. Naive B cells were isolated and cultured for 24 h in IL-4/anti-CD40 or without exogenous cytokines (unstimulated) before nonquantitative single cell RT-PCR analysis for and 1–4 GLT. GLT PCR products were separated by electrophoresis, blotted onto a nylon membrane, and probed together using a probe specific to all GLT. PCR products were subjected to electrophoresis, blotted, and probed with an -specific probe separately. After autoradiography, the data were arranged to display the and GLT results for each cell together. Due to the high number of cycles used, additional nonspecific bands were occasionally seen in the PCRs. The correctly sized band, representing the mature GLT product, is labeled with an arrow. The results shown are representative of three different experiments, with the exception of 3 detection in all IL-4/anti-CD40-stimulated cells. It is possible that this unusually high 3 signal arose due to patient variability; 3 is constitutively expressed at a high level. The possibility of contamination can be ruled out due to the absence of bands in the negative control lanes.

View larger version (27K): [in this window] [in a new window]   FIGURE 4. Semiquantitative RT-PCR analysis of GLT and GAPDH expression in single naive B cells. Naive tonsillar B cells were isolated as previously described and cultured for 24 h in IL-4/anti-CD40 or without exogenous cytokines (unstimulated). Single cell RT-PCR was conducted as described in Fig. 3, except that samples were taken from the first round of PCR amplification after 14, 18, 22, and 26 cycles. Samples from each of these cycle numbers were then subjected to the second round of PCR, using nested primers, for 18 cycles and analyzed as previously discussed. A, Semiquantitative analysis of GLT expression. B, Semiquantitative analysis of GAPDH expression in the same cells. C, Detection of known amounts of a cloned GLT template. GLT template of a known concentration was serially diluted in tRNA (0.5µg/ml) to give samples of known copy numbers. These samples were subjected to two rounds of PCR amplification, each of 18 cycles, thus representing the point at which GLT are detected in the unstimulated cells.

Multiple GLT in single naive B cells

The nonquantitative single cell RT-PCR method used to study the distribution of GLT in single cells was extended to the four subclasses. Fig. 5 shows an analysis of and GLT in single naive tonsil B cells cultured for 24 h with IL-4 and anti-CD40 or without exogenous cytokines; similar patterns of expression have been seen with IL-10 and IFN- stimulation (results not shown). The results show that some B cells, whether stimulated or unstimulated, coexpress GLT and GLT in some of the same cells. All combinations of GLT isotypes have been seen, with cells expressing only one isotype, or 1–4, a combination of two or more GLT, or a combination of GLT and GLT. As with GLT, transcripts were present in a large proportion of unstimulated cells, whereas stimulation with any of the cytokines had little effect on this proportion.

The expression of multiple isotypes in the majority of cells reveals that a large stretch of the H chain locus encompassing the 3 to genes is in a constitutively open chromatin conformation in these cells.

Comparison of GLT and globin gene expression in single B cells, T cells, and erythroid cells

To confirm the significance of the levels of constitutive GLT expression detected in the majority of single B cells, we compared these findings with a semiquantitative analysis of GLT and globin expression in two other cell types, primary human T cells and the erythroleukemic cell line K562 (31, 38). Fig. 6 shows that whereas globin transcripts were readily detected in single K562 cells, they were not detected in single human B or T cells: globin transcripts were detected after 20 second-round PCR cycles in K562 cells, but no transcripts were detected after 32 cycles in the single human B cells (data not shown). Conversely, GLT were detected in the majority of single B cells, but were undetectable in the K562 cells: GLT were detectable after 24 first round cycles in B cells but were undetectable after 32 cycles in the K562 cells (data not shown). Perhaps surprisingly, GLT were also readily detectable in primary human T cells at similar cycle numbers to the unstimulated B cells. This data validate our conclusion that transcripts are not detectable from fully repressed heterochromatic genes, globin in B/T cells and GLT in K562 cells. In contrast, the H chain locus is in a constitutively open state in B and T cells and is transcribed, albeit at a low level, in these cell types. However, as demonstrated above, transcription is markedly increased from this locus upon IL-4 stimulation.

View larger version (58K): [in this window] [in a new window]   FIGURE 6. Comparison of GLT and globin gene expression in single B cells, T cells, and erythroid cells. Naive human B cells, human T cells, and the erythroleukemic cell line, K562, were isolated as previously described. Single-cell RT-PCR was conducted immediately after cell isolation as described in Fig. 4. Samples were taken from the first round of PCR amplification after 14, 18, 22, and 26 cycles. Samples from each of these cycle numbers were then subjected to the second round of PCR, using the nested primers previously discussed. All cDNA and PCRs were conducted concurrently, using the same reaction mix, for each cell line. Semiquantitative analysis of globin expression in these cells is shown after 14 first-round and 24 second-round PCR cycles. Semiquantitative analysis of GLT expression is shown in the same cells after identical cycle numbers. In each case, GAPDH reactions were performed to validate the cDNA reaction.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The evidence from single cell analysis was central to this conclusion. The same approach has previously been used on hemopoietic stem cells to demonstrate transcription of genes encoding transcription factors specific for all subsequent lineages (32). Chromatin accessibility has also been invoked in allelic exclusion in B cells, but Singh et al. (40) showed, by analysis of single cells, that the locus is biallelically expressed in single immature B cells despite the operation of allelic exclusion mechanisms. From this, the authors concluded that selective unfolding of chromatin higher order structure is not involved in biallelic exclusion.

Constitutive transcription of multiple germline genes

Although we have found that 40% of the single naive tonsil and peripheral blood B cells express one or more germline genes, this is almost certainly a lower limit. Thus, 1) cells may express GLT below the level of detection or the RT-PCR may have failed; 2) cells that expressed a GLT, but no perceptible GAPDH, were excluded from the count, even though the lack of GAPDH expression was probably due to a failure of that reaction; and 3) germline gene transcription is restricted to the G₁/S boundary of the cell cycle (41), and some cells would probably have been in other phases.

The observation of constitutive GLT expression in human IgD⁺ tonsil B cells is not unprecedented, although there have been considerable differences between the results from different laboratories. Fujieda et al. (28) and Kitani and Strober (42) reported constitutive 1 and 2, but no 3 or 4, GLT, respectively, in freshly isolated IgD⁺ human tonsil B cells and B cells cultured for 48 h; GLT were not examined. Gauchat et al. (43) and Jumper et al. (27) did not detect constitutive GLT in unstimulated IgD⁺ tonsil B cells after 5–8 days of incubation, but GLT may have been degraded by that time. Revy et al. (44) did not detect constitutive GLT in freshly isolated PBMC, but the B cells would have constituted only some 5% of the total number of cells, so that germline gene transcription may have escaped detection. Cerutti et al. (30) detected constitutive GLT in only the 15% of tonsillar IgD⁺ B cell populations that were CD38⁺, identifying them as germinal center founder cells.

Although it is likely that a part of our detected GLT signal probably derived from the founder cell population, it could not account for the GLT signals we found in >40% of unstimulated naive B cells. Our results demonstrate that in unstimulated cells GLT are expressed at a significant level (leading to a steady state concentration of >50 copies/cell). These results also crucially eliminate the possibility that the levels of constitutive GLT expression detected in the population reflects a small number of previously activated cells expressing the gene at high levels.

The significance of this low level of constitutive transcription was demonstrated by comparison with a gene that is in a heterochromatic state in B cells. Fig. 6 demonstrates that globin transcripts are undetectable in single B cells (and T cells) while being readily detectable in the K562 erythroleukemia cell line, thus validating the globin PCR detection method. This result would be expected from a gene that is in a heterochromatic conformation in the B cells and therefore is unable to be transcribed. In contrast, although GLT are undetectable in the K562 cell line, they are detectable at similar levels in both primary B- and T cells. These results support our conclusion that the locus (and, in fact, the entire Ig H chain locus) is in an open chromatin conformation in B cells and demonstrate that it is also open in the T cell lineage. However, the absence of an GLT signal in the erythroid K562 cells demonstrates that this result is not a consequence of the GLT locus being inherently "leaky" in all cells, but is indicative of the open nature of the locus in the lymphoid lineage.

Effects of cytokine stimulation

We have shown, in agreement with previous authors (27, 28, 29, 30), that IL-4 significantly enhances the expression of , 2, 3, and 4 GLT within 24 h and the secretion of the corresponding proteins after 10-day incubation. Our semiquantitative data also demonstrate that 1) IL-4 up-regulates the expression of and GLT to levels equal to or greater than those engendered by other cytokines; and 2) IL-4 up-regulates all germline genes to similar extents, but the gene to a level 10 times higher than the genes; nevertheless, 3) the level of secretion of IgE stimulated by IL-4 is 10 times lower than that of IgG, whereas 4) the levels of IgG secretion, excluding IgG4, stimulated by IL-4 or IL-10 plus anti-CD40, are comparable. These results demonstrate significant uncoupling of GLT and protein expression, underscoring the complexity of isotype determination (2, 4).

Most importantly, our results confirm the finding that in human B cells IL-4 is not specific for (or 4) germline gene transcription and does not suppress 1, 2, or 3. This is contrary to the reported situation in murine B cells (19, 20, 21, 22). Conversely, IFN- did not inhibit GLT in single human B cells. The apparent difference between murine and human B cells may be a fundamental one or may merely reflect the more sensitive methods that became available for the more recent studies of human B cells. Failure to observe GLT in murine B cells stimulated by IL-4 may also have been due to the time elapsed between observation and initiation of the cultures (19), given the short half-lives of GLT. We have observed that the levels of GLT diminish with time, whereas GLT persist for longer (unpublished observations), presumably due to sequential switching and decay of the previously synthesized transcripts.

Our comparative analysis of unstimulated and IL-4-stimulated single cells allowed us to distinguish between two possible modes of IL-4 stimulation, i.e., an increase in 1) the number of cells active in germline gene transcription, or 2) the number of GLT transcribed in the constitutively active cells. In fact, there was little scope for an increase in cell numbers that would account for the stimulation of and 4 germline gene transcription by 2 orders of magnitude, because >40% of cells already express the GLT before stimulation. Indeed, no significant increase in the proportion of cells expressing any of the isotypes, with any of the cytokine combinations, was seen. These results demonstrate that IL-4 acts by increasing the level of GLT expression in cells that are already transcribing the locus, albeit at the lower basal level.

Alternative models of isotype determination

If chromatin accessibility is not involved in isotype determination, what could be the mechanism? The level of activity of individual genes in an open chromatin domain can be modulated by local covalent modifications, in particular, histone methylation and acetylation, and DNA methylation, as well as by chromatin remodeling causing nucleosome repositioning. In studies in mouse B cells, histone acetylation per se has been shown to be unnecessary for CSR (45). Demethylation of cytosine has been shown to accompany VDJ recombination and allelic exclusion (46, 47), and demethylation of H chain germline genes undergoing germline gene transcription has been reported in studies of murine B cells (18, 48, 49). The effect of methylation status on the susceptibility of cytosines to the action of AID and CSR has yet to be determined.

The fact that CSR to IgE predominantly occurs by sequential switching is likely to be important in isotype determination. Tangye et al. (50) have shown that class switching to IgE in both murine and human B cells is related to cell division; IgE requires more cycles than IgG in class switching stimulated by IL-4. This would point to a complex model of isotype regulation, incorporating different probabilities of CSR to the C and C loci, along with sequential CSR (delaying switching to IgE). DNA binding factors are thought to control the relative frequency of CSR at defined switch regions (51) and may be influenced by cytokines. The proximity of the germline genes to one another in the nucleus (52, 53), DNA sequence (7, 54), and the conformation of specific switch regions induced by transcription (7), are potential factors in determining the frequency of CSR. Competition between signals inducing cell proliferation (e.g., IL-4, IL-10 (55), and CD40L) and cell differentiation (e.g., IL-10 (56)) may determine when cells cease dividing and undergo apoptosis or differentiate to plasma cells.

Although IL-4 is not required to unfold the higher order structure of the germline gene, it may nevertheless have multiple effects on recombination and cell differentiation. IL-4 and IL-13 are the only cytokines that stimulate the expression of GLT. However, IL-4 also up-regulates the expression of AID (44) and CD23 (57). AID is required for CSR (5, 6), and CD23 rescues B cells from apoptosis, thereby enhancing cell proliferation to allow sequential switching (50) and plasma cell differentiation (58). Additional analysis of the interplay between all relevant factors is required for a full understanding of isotype determination.

	Acknowledgments

 
Tonsil material was kindly collected by the GKT Health Care Trust Ear, Nose, and Throat Department and surgery staff.

	Footnotes

 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ This work was supported by British Lung Foundation Project Grant P007 and Wellcome Trust Program Grant 056225/Z.98/Z.

² Address correspondence and reprint requests to Dr. Hannah J. Gould, The Randall Center, King’s College London, New Hunt’s House, London, U.K. SE1 1UL. E-mail address: hannah.gould{at}kcl.ac.uk

³ Abbreviations used in this paper used: CSR, class switch recombination; AID, activation-induced cytidine deaminase; GLT, germline gene transcript.

Received for publication September 23, 2003. Accepted for publication August 3, 2004.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Max, E. E.. 1998. Immunoglobulins: molecular genetics. W. E. Paul, ed. Fundamentals of Immunology 111. Lippincott, Philadelphia.
2. Gould, H. J., B. J. Sutton, A. J. Beavil, R. L. Beavil, N. McCloskey, H. A. Coker, D. Fear, L. Smurthwaite. 2003. The biology of IgE and the basis of allergic disease. Annu. Rev. Immunol. 21:579.[Medline]
3. Stavnezer, J.. 2000. Molecular processes that regulate class switching. Curr. Top. Microbiol. Immunol. 245:127.[Medline]
4. Gould, H. J., R. L. Beavil, D. Vercelli. 2000. IgE isotype determination: epsilon-germline gene transcription, DNA recombination and B-cell differentiation. Br. Med. Bull. 56:908.[Abstract/Free Full Text]
5. Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, T. Honjo. 2000. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102:553.[Medline]
6. Okazaki, I., K. Yoshikawa, K. Kinoshita, M. Muramatsu, H. Nagaoka, T. Honjo. 2003. Activation-induced cytidine deaminase links class switch recombination and somatic hypermutation. Ann. NY Acad. Sci. 987:1.[Abstract/Free Full Text]
7. Yu, K., M. R. Lieber. 2003. Nucleic acid structures and enzymes in the immunoglobulin class switch recombination mechanism. DNA Repair 2:1163.[Medline]
8. Neuberger, M. S., R. S. Harris, J. Di Noia, S. K. Petersen-Mahrt. 2003. Immunity through DNA deamination. Trends Biochem. Sci. 28:305.[Medline]
9. Fugmann, S. D., D. G. Schatz. 2003. RNA AIDS DNA. Nat. Immunol. 4:429.[Medline]
10. Yu, K., F. Chedin, C. L. Hsieh, T. E. Wilson, M. R. Lieber. 2003. R-loops at immunoglobulin class switch regions in the chromosomes of stimulated B cells. Nat. Immunol. 4:442.[Medline]
11. Ramiro, A. R., P. Stavropoulos, M. Jankovic, M. C. Nussenzweig. 2003. Transcription enhances AID-mediated cytidine deamination by exposing single-stranded DNA on the nontemplate strand. Nat. Immunol. 4:452.[Medline]
12. Chen, X., K. Kinoshita, T. Honjo. 2001. Variable deletion and duplication at recombination junction ends: implication for staggered double-strand cleavage in class-switch recombination. Proc. Natl. Acad. Sci. USA 98:13860.[Abstract/Free Full Text]
13. Hein, K., M. G. Lorenz, G. Siebenkotten, K. Petry, R. Christine, A. Radbruch. 1998. Processing of switch transcripts is required for targeting of antibody class switch recombination. J. Exp. Med. 188:2369.[Abstract/Free Full Text]
14. Kinoshita, K., T. Honjo. 2000. Unique and unprecedented recombination mechanisms in class switching. Curr. Opin. Immunol. 12:195.[Medline]
15. Calame, K. L.. 2001. Plasma cells: finding new light at the end of B cell development. Nat. Immunol. 2:1103.[Medline]
16. Calame, K. L., K. I. Lin, C. Tunyaplin. 2003. Regulatory mechanisms that determine the development and function of plasma cells. Annu. Rev. Immunol. 21:205.[Medline]
17. Blackwell, T. K., M. W. Moore, G. D. Yancopoulos, H. Suh, S. Lutzker, E. Selsing, F. W. Alt. 1986. Recombination between immunoglobulin variable region gene segments is enhanced by transcription. Nature 324:585.[Medline]
18. Stavnezer-Nordgren, J., S. Sirlin. 1986. Specificity of immunoglobulin heavy chain switch correlates with activity of germline heavy chain genes prior to switching. EMBO J. 5:95.[Medline]
19. Lutzker, S., P. Rothman, R. Pollock, R. Coffman, F. W. Alt. 1988. Mitogen- and IL-4-regulated expression of germ-line Ig 2b transcripts: evidence for directed heavy chain class switching. Cell 53:177.[Medline]
20. Rothman, P., S. Lutzker, W. Cook, R. Coffman, F. W. Alt. 1988. Mitogen plus interleukin 4 induction of C epsilon transcripts in B lymphoid cells. J. Exp. Med. 168:2385.[Abstract/Free Full Text]
21. Severinson, E., C. Fernandez, J. Stavnezer. 1990. Induction of germ-line immunoglobulin heavy chain transcripts by mitogens and interleukins prior to switch recombination. Eur. J. Immunol. 20:1079.[Medline]
22. Berton, M. T., J. W. Uhr, E. S. Vitetta. 1989. Synthesis of germ-line 1 immunoglobulin heavy-chain transcripts in resting B cells: induction by interleukin 4 and inhibition by interferon . Proc. Natl. Acad. Sci. USA 86:2829.[Abstract/Free Full Text]
23. Coffman, R. L., D. A. Lebman, P. Rothman. 1993. Mechanism and regulation of immunoglobulin isotype switching. Adv. Immunol. 54:229.[Medline]
24. Lorenz, M., A. Radbruch. 1996. Developmental and molecular regulation of immunoglobulin class switch recombination. Curr. Top. Microbiol. Immunol. 217:151.[Medline]
25. Zhang, K.. 2003. Accessibility control and machinery of immunoglobulin class switch recombination. J. Leukocyte Biol. 73:323.[Abstract/Free Full Text]
26. Splawski, J. B., S. M. Fu, P. E. Lipsky. 1993. Immunoregulatory role of CD40 in human B cell differentiation. J. Immunol. 150:1276.[Abstract]
27. Jumper, M. D., J. B. Splawski, P. E. Lipsky, K. Meek. 1994. Ligation of CD40 induces sterile transcripts of multiple Ig H chain isotypes in human B cells. J. Immunol. 152:438.[Abstract]
28. Fujieda, S., K. Zhang, A. Saxon. 1995. IL-4 plus CD40 monoclonal antibody induces human B cells subclass-specific isotype switch: switching to 1, 3, and 4, but not 2. J. Immunol. 155:2318.[Abstract]
29. Fujieda, S., Y. Q. Lin, A. Saxon, K. Zhang. 1996. Multiple types of chimeric germ-line Ig heavy chain transcripts in human B cells: evidence for trans-splicing of human Ig RNA. J. Immunol. 157:3450.[Abstract]
30. Cerutti, A., H. Zan, E. C. Kim, S. Shah, E. J. Schattner, A. Schaffer, P. Casali. 2002. Ongoing in vivo immunoglobulin class switch DNA recombination in chronic lymphocytic leukemia B cells. J. Immunol. 169:6594.[Abstract/Free Full Text]
31. Miller, C. W., K. Young, D. Dumenil, B. P. Alter, J. M. Schofield, A. Bank. 1984. Specific globin mRNAs in human erythroleukemia (K562) cells. Blood 63:195.[Abstract/Free Full Text]
32. Yssel, H., J. E. de Vries, M. Koken, W. Van Blitterswijk, H. Spits. 1984. Serum-free medium for generation and propagation of functional human cytotoxic and helper T cell clones. J. Immunol. Methods 72:219.[Medline]
33. Hu, M., D. Krause, M. Greaves, S. Sharkis, M. Dexter, C. Heyworth, T. Enver. 1997. Multilineage gene expression precedes commitment in the hemopoietic system. Genes Dev. 11:774.[Abstract/Free Full Text]
34. Testi, R., D. D’Ambrosio, R. De Maria, A. Santoni. 1994. The CD69 receptor: a multipurpose cell-surface trigger for hematopoietic cells. Immunol. Today 15:479.[Medline]
35. Agematsu, K., S. Hokibara, H. Nagumo, A. Komiyama. 2000. CD27: a memory B-cell marker. Immunol. Today 21:204.[Medline]
36. Kenter, A. L.. 2003. Class-switch recombination: after the dawn of AID. Curr. Opin. Immunol. 15:190.[Medline]
37. Flanagan, J. G., T. H. Rabbitts. 1982. The sequence of a human immunoglobulin epsilon heavy chain constant region gene, and evidence for three non-allelic genes. EMBO J. 1:655.[Medline]
38. Tuan, D., I. M. London. 1984. Mapping of DNase I-hypersensitive sites in the upstream DNA of human embryonic epsilon-globin gene in K562 leukemia cells. Proc. Natl. Acad. Sci. USA 81:2718.[Abstract/Free Full Text]
39. Mostoslavsky, R., F. W. Alt, C. H. Bassing. 2003. Chromatin dynamics and locus accessibility in the immune system. Nat. Immunol. 4:603.[Medline]
40. Singh, N., Y. Bergman, H. Cedar, A. Chess. 2003. Biallelic germline transcription at the immunoglobulin locus. J. Exp. Med. 197:743.[Abstract/Free Full Text]
41. Lundgren, M., L. Strom, L. O. Bergquist, S. Skog, T. Heiden, J. Stavnezer, E. Severinson. 1995. Cell cycle regulation of immunoglobulin class switch recombination and germ-line transcription: potential role of Ets family members. Eur. J. Immunol. 25:2042.[Medline]
42. Kitani, A., W. Strober. 1993. Regulation of C subclass germ-line transcripts in human peripheral blood B cells. J. Immunol. 151:3478.[Abstract]
43. Gauchat, J. F., D. A. Lebman, R. L. Coffman, H. Gascan, J. E. de Vries. 1990. Structure and expression of germline transcripts in human B cells induced by interleukin 4 to switch to IgE production. J. Exp. Med. 172:463.[Abstract/Free Full Text]
44. Revy, P., T. Muto, Y. Levy, F. Geissmann, A. Plebani, O. Sanal, N. Catalan, M. Forveille, R. Dufourcq-Labelouse, A. Gennery, et al 2000. Activation-induced cytidine deaminase (AID) deficiency causes the autosomal recessive form of the hyper-IgM syndrome (HIGM2). Cell 102:565.[Medline]
45. Nambu, Y., M. Sugai, H. Gonda, C. G. Lee, T. Katakai, Y. Agata, Y. Yokota, A. Shimizu. 2003. Transcription-coupled events associating with immunoglobulin switch region chromatin. Science 302:2137.[Abstract/Free Full Text]
46. Krangel, M. S.. 2001. V(D)J recombination becomes accessible. J. Exp. Med. 193:F27.
47. Bergman, Y., A. Fisher, H. Cedar. 2003. Epigenetic mechanisms that regulate antigen receptor gene expression. Curr. Opin. Immunol. 15:176.[Medline]
48. Burger, C., A. Radbruch. 1992. Demethylation of the constant region genes of immunoglobulins reflects the differentiation state of the B cell. Mol. Immunol. 29:1105.[Medline]
49. Kuwabara, N., N. Kondo, O. Fukutomi, H. Fujii, T. Orii. 1995. Methylation patterns of I epsilon region in B cells stimulated with interleukin 4 and Epstein-Barr virus in patients with a high level of serum IgE. Eur. J. Immunogenet. 22:265.[Medline]
50. Tangye, S. G., A. Ferguson, D. T. Avery, C. S. Ma, P. D. Hodgkin. 2002. Isotype switching by human B cells is division-associated and regulated by cytokines. J. Immunol. 169:4298.[Abstract/Free Full Text]