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 Ford, G. S. Articles by Covey, L. R. Search for Related Content PubMed PubMed Citation Articles by Ford, G. S. Articles by Covey, L. R.

CD40 Ligand Exerts Differential Effects on the Expression of I Transcripts in Subclones of an IgM⁺ Human B Cell Lymphoma Line¹

Division of Life Sciences, Rutgers, The State University of New Jersey, Piscataway, NJ 08855

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The CD40:CD40 ligand (CD40L) interaction plays a critical role in T cell-dependent isotype switching. To elucidate the role of CD40 signaling in the activation of germline transcription and as an extension, in targeting C regions for isotype switching, an IgM⁺ Burkitt lymphoma cell line (Ramos 2G6) was assayed for the up-regulation of germline transcripts after CD40L stimulation. Independent Ramos 2G6 subclones that either expressed (I⁺) or did not express (I^-) basal levels of I transcripts were assessed for their transcriptional response to CD40L signaling by contact with either a Jurkat T cell line (D1.1) or a transfected CD40L-expressing epithelial cell line (293/CD40L) in the presence or absence of IL-4. Both I^- and I⁺ Ramos 2G6 subclones cultured with IL-4 and CD40L markedly up-regulated germline transcription predominantly from the 1, 2, and 3 subclasses over levels obtained with IL-4 alone. In addition, these two signals were required to obtain de novo switch recombination. However, incubation with CD40L alone resulted in a substantial increase in germline transcription only in the I⁺ and not the I^- subclones. Observed basal transcription at the 1 locus also correlated with the ability of not only the 1 locus, but also the 2 and 3 loci, to up-regulate germline transcripts in response to CD40 signaling. These data are consistent with CD40:CD40L contact up-regulating germline transcription only after the B cell has received a signal that alters the transcriptional state of the heavy chain locus.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The differentiation of B cells in response to T cell-dependent (TD)³ Ags requires cognate interactions between CD40 expressed on B cells and CD40 ligand (CD40L) expressed on activated CD4⁺ T cells (reviewed in 1 . The consequences of this interaction are that Ag-activated B cells undergo proliferation, homotypic adhesion, and class switch recombination, or isotype switching (2, 3, 4, 5, 6), to produce new clonatypic Abs that differ with respect to their C region effector functions, but still retain their unique specificity with regard to Ag binding.

In addition to the requirement for CD40L in isotype switching, cytokines are known to be involved in regulating the isotype profile in response to an antigenic challenge (for review, see 7 . It is proposed that cytokines function at one level by regulating the accessibility of the different C_H regions to switch factors, and that this process is correlated with chromatin changes that include increased DNase I hypersensitivity (8, 9) and hypomethylation (10) of specific C_H regions, as well as the activation of transcription from sites upstream of the C_H-associated switch (S) regions (11, 12, 13, 14, 15, 16). The appearance of small C_H-specific RNA transcripts, termed germline or sterile transcripts, is widely seen as an obligatory event preceding isotype switching in both mice (11, 12, 13, 14, 15, 17, 18, 19) and humans (16). Although germline transcription can be directly regulated by cytokines, soluble factors alone are clearly insufficient to induce significant TD class switching (20, 21, 22, 23, 24). This process requires an additional signal provided by contact through CD40L expressed on activated CD4⁺ T cells (reviewed in Refs. 25 and 26).

The critical importance of the CD40:CD40L interaction for T-dependent humoral immune responses has been demonstrated by an absence of switch recombination and deficient humoral immunity in animals lacking functional CD40 or CD40L (human X-linked hyper IgM syndrome (HIGMX-1) (27, 28, 29, 30, 31, 32); and CD40 (33)- and CD40L (34)-deficient mice). CD40 engagement mediates several cellular events, including the activation of multiple protein kinases and the specific phosphorylation of phospholipase C-2 and phosphoinositide-3' kinases (35, 36, 37, 38). In addition, nuclear factor-B and other rel family members are induced in response to CD40 ligation (37, 39, 40, 41). However, the relationship between these initial signaling events and downstream processes, such as switch recombination, is poorly understood. With respect to isotype switching, CD40 signaling appears to result both in the activation of the switch recombinase machinery and in chromatin changes (i.e., germline transcription) that may impart recombinational accessibility to the C_H locus (2, 3, 5, 42, 43, 44, 45, 46).

It has been widely reported that CD40 signaling in the presence of IL-4 augments the level of germline transcription over the level obtained with IL-4 alone (43, 44, 47). However, it is unclear whether contact solely through CD40 can induce germline transcription. Studies conducted with murine B cells suggest that membranes from activated Th cells or CD40L expressed in Sf9 cells can induce germline transcription in the absence of cytokines (48, 49). However, other investigators found that activated Th membranes were insufficient to induce germline transcription (50). Studies using human B cells have also produced conflicting results. While specific studies revealed that multiple classes of germline transcripts in purified peripheral B cells are induced in response to CD40 engagement (44, 51), others found that germline transcripts were not induced in response to signaling through CD40 (43, 46, 47). The results in many of these studies could be attributed to the different methods used to isolate B cells and also to the different sensitivities of the methods used for detecting germline transcripts (PCR vs RNase protection). Therefore, it remains unclear whether CD40 signaling mediates the induction of germline transcription or whether transcriptional targeting of specific C_H loci is primarily a consequence of cytokine signaling.

To examine this question more fully, we have conducted experiments using an IL-4-responsive Burkitt lymphoma cell line (Ramos 2G6) (52) that undergoes C_H transcriptional activation and limited class switching in response to TD signals (43, 53, 54). Isolated subclones that differed in their basal expression of I transcripts (I⁺ and I^-) were used in parallel to examine the effect CD40 stimulation has on the transcriptional activation of the different C genes.

We report in this work that CD40 signaling differentially regulates the expression of I transcripts in our defined model system. CD40L contact, in the absence of IL-4, was found to up-regulate I transcription in the I⁺ subclones. However, under identical conditions, marginal to no I transcription was detected in the I^- population. Finally, in confirmation of others’ findings (43, 44, 47, 49, 55, 56), measurable switch recombination was observed only after stimulation with both IL-4 and CD40L, conditions that produced the highest level of I expression in both I^- and I⁺ subclones. Our results support a role for CD40L in up-regulating I expression only after the heavy chain locus has become transcriptionally active, and therefore argue against CD40L having a primary role in the selection of C_H genes involved in isotype switching.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cell lines

The Jurkat clones D1.1 (CD40L positive) and B2.7 (CD40L negative) have been described previously (57, 58). The human B cell lymphoma clone Ramos 2G6.4CN3F10 (Ramos 2G6) is an IL-4-responsive subclone of RA-1 (59) and has been previously described (52). B and T cell lines were maintained in RPMI 1640 supplemented with 10% FBS (Atlanta Biologics, Atlanta, GA), 100 U/ml penicillin, 100 µg/ml streptomycin, and 2 mM glutamine. 293 cells are derived from a primary human embryonal kidney cell transformed by adenovirus 5 and are available from American Type Culture Collection (Rockville, MD). The 293/CD40L line was constructed by the stable transfection of pCT-BAM into 293 cells, as previously described (54). Untransfected 293 and 293/CD40L cells were maintained in DMEM/F12 media with 10% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, and 2 mM glutamine.

Abs and cytofluorographic analysis

For evaluation of cell surface Ig expression, FITC-labeled goat F(ab')₂ anti-human IgM and anti-human IgG were purchased from Southern Biotechnology (Birmingham, AL). R-phycoerythrin (PE)-conjugated mouse anti-human CD20 mAb was purchased from PharMingen Corp. (San Diego, CA). Analysis of surface CD40L expression was conducted using the anti-CD40L mAb, 5c8, previously described by Lederman et al. (58). Approximately 2 x 10⁵ cells were incubated with saturating concentrations of the indicated mAbs for 30 min at 4°C. After washing the cells once with 3.5 ml of FACS wash (MEM + 12 mM HEPES, pH 7.2, 0.2% NaHC02), cells were resuspended in 150 µl FACS wash and fluorescence intensity measured using an Epics Profile II flow cytometer (Coulter Electronics, Hialeah, FL).

Mitomycin-C treatment

Jurkat D1.1 and B2.7, 293 cells, or 293/CD40L cells (10⁷/ml) were incubated with 50 µg/ml of mitomycin-C (Sigma Chemical Co., St. Louis, MO) for 45 min at 37°C/5% CO₂ to inhibit cell division. The treated cells were washed three times with PBS and allowed to incubate for 1 h at 37°C/5% CO₂. Cells were again washed three times with PBS and resuspended in RPMI/10% FCS for use in T cell/B cell experiments.

T/B cocultures

To examine C transcription, 2 x 10⁵ Ramos 2G6 B cells were cultured for 6 days in 1 ml of RPMI/10% FBS, either alone or with mitomycin-C-treated Jurkat cells (D1.1 or B2.7) at a ratio of 2.5 to 5 T cells:1 B cell. Human rIL-4 (Life Technologies, Gaithersburg, MD) was added to replica wells at a concentration of 400 U/ml. Cocultures were established with 293 or 293/CD40L cells using the same conditions.

RNA isolation and PCR strategy

After the indicated incubation period, cells were harvested and RNA isolated using Trizol reagent (Life Technologies). cDNA was prepared by transcribing 2 µg of total RNA from the T/B cocultures in a 25-µl vol containing 200 µM of each dNTP, 10 pmol of random primer (Promega Corp., Madison, WI), 20 U RNasin (Promega Corp.), and 50 U AMV-RT (avian myeloblastosis virus reverse transcriptase) (Promega Corp.). After incubation at 37°C for 1 h, and heat inactivation for 10 min at 68°C, 5 µl was used in a 100-µl PCR reaction containing 200 µM of each dNTP, and 400 ng of 5' and 3' primers. Reactions were heated to 72°C for 3 min, after which 2.5 U of Taq polymerase (Promega Corp.) was added to each tube. The cycling conditions were as follows: 1.5 min at 94°C, 1 min at 55°C, and 1 min at 72°C for 30 cycles.

Oligonucleotides for RT-PCR

The PCR assay used to amplify total RNA from B/T cocultures was developed and described by Jumper et al. (44). The sequences of oligos used to amplify germline transcripts are found in the I region (5'-I), 5'-gccctcctctcagccaggacc-3') and in the second exon of the C region (3'-C_H2 C, 5'-tccttgggttttggggggaa-3'). These oligos amplify I transcripts from all four subclasses. Subclass-specific hinge probes used to identify specific I and VDJ-C transcripts are as follows: hinge 1, 5'-aaatcttgtgacaaaactcaca-3'; hinge 2, 5'-gttgtgtcgactgcccacc-3'; hinge 3, 5'-tgcccacggtgcccagag-3'; and hinge 4, 5'-tgagtccaaatatggtcccc-3'. The specific size of each amplified product was based on previously reported sequences for the different subclasses (60, 61, 62, 63, 64). For synthesis of VDJ-C and VDJ-Cµ PCR products, the following primers were used: VDJ-C, (5'-J_H) 5'-acc^c/_atggtcaccgyctcctca-3' and (3'-C_H2 C) 5'-tccttgggtttgggggaa-3'; VDJ-Cµ, (5'-V_H1.3) 5'-ggacacggctgtgtatta-3' and (3'-Cµ1) 5'-gggaattcaaggaagtcctgtgcgag-3'. Semiquantitative PCR was conducted using conditions described above, but with 10-fold and 100-fold less cDNA. Primers specific for I and Cµ were included in each reaction. Quantification of the I signal was determined by dividing the I signal at each point by the Cµ signal. We did not detect a difference in the expression of Cµ under the different conditions of our assay. This was established by determining the absolute numerical values of the Cµ signal by phosphor image analysis and quantitation.

S1 nuclease protection assays

T cell/B cell cultures were established as indicated above and RNA isolated using the Trizol method. The probe used in these studies has been described previously (43), and is a uniformly labeled RNA probe produced by in vitro transcription of a 440-bp SacI fragment derived from the I3-C3 cDNA. A quantity amounting to 200,000 cpm of labeled probe was incubated together with 15 µg of total RNA for 20 h in 80% formamide, 40 mM PIPES, 1 mM EDTA, and 400 mM NaCl at 60°C. The reaction mixture was then digested with 100 U of S₁ nuclease (Promega Corp.) for 1 h at 37°C, followed by electrophoresis on a 5% acrylamide/urea gel to resolve protected bands.

Southern blotting and hinge region-specific hybridization

PCR products were separated on 1.5% agarose gels and transferred to nylon membranes for Southern blotting. In experiments in which C subclass specificity was determined, four replicate gels were run. Blots were pre-hybridized for 1 h, followed by overnight hybridization at the appropriate temperature with 1 x 10⁷ cpm/ml of kinased hinge region probe. The following temperatures were used for prehybridization and hybridization: 1, 52°C; 2 to 4, 56°C; and 5' J_H, 56°C. Blots were washed once at room temperature and twice at the probe-specific temperature in 2x SSC + 0.5% SDS. To quantitate signals in experiments conducted semiquantitatively, filters were exposed to a phosphor screen and analyzed using a Storm Imaging System (Molecular Dynamics, Sunnyvale, CA).

Isolation and identification of reciprocal switch circle products

Isolation of switched circle products: 2 x 10⁵ Ramos B cells from the A/G8 (I^-) subclone were cultured either alone, or with 0.5 x 10⁶ mitomycin-treated 293/CD40L cells with or without 400 U of human rIL-4. Mitomycin treatment and coculture conditions were identical to those described above. After 3 days, cultures were refed with RPMI complete and rIL-4, where appropriate.

After 6 days, supercoiled circular DNA molecules were isolated using an alkaline lysis protocol described by Zhang et al. (65). Briefly, cell pellets were resuspended in 400 µl alkaline lysis buffer (50 mM NaCl, 2 mM EDTA, and 1% SDS, pH 12.4), vortexed for 5 min, and incubated at 30°C for 1 h. After addition of 20 µl 1 M Tris-HCl, 7.4 and 40 µl 5 M NaCl, lysates were incubated with proteinase K (100 µg/ml final concentration) for 30 min at 37°C. After phenol/chloroform extractions and ethanol precipitation, DNA pellets were resuspended in 50 µl Tris-EDTA (pH7.4) and digested for 1 h with EcoRI and RNase A before PCR amplification. As a positive control for amplification, we used 10 pg of the Sµ-S1 plasmid described by Malisan et al. (66).

PCR amplification

Samples were subjected to two rounds of amplification using a modified protocol of Fujieda et al. (46) and Malisan et al. (66). Briefly, 0.5 µl of digested circular DNA was amplified in a 50-µl reaction containing 50 mM KCl, 10 mM Tris-HCl, pH 9, 1.5 mM MgCl₂, 0.1% Triton X-100, 200 µM dNTP, 5% DMSO, 0.5 U Taq DNA polymerase (Promega Corp.), and 100 nM of each primer. Samples were incubated initially at 95°C for 10 min, 60°C for 10 min (at which point the enzyme was added), and 72°C for 10 min. These steps were followed by 40 cycles of: 94°C for 1 min, 65°C for 1 min, and 72°C for 2 min. The second round of amplification was conducted with 5 µl of the first round products using the same conditions and cycling times as described above. The primers used in the first round amplification were as follows: M₁, 5'-ggtgagtgtgatggggaacgcagtgta-3', corresponding to nucleotides 3870–3844 of Sµ; and G₁, 5'-gggcttccaagccaacagggcaggaca-3', corresponding to nucleotides 1859–1885 in the S4 region (67). The primers used in the second round were: M₁ (described above) and G3, 5'-aagagtccagggaggcccagaaaggcccag-3', corresponding to nt 1193–1222 in the S region (68).

Ten microliters of the second round PCR products were separated on a 1% agarose gel and transferred to nylon membrane. Reciprocal switch products were identified using a random primed S probe described by Malisan et al. (66). This probe includes the nucleotides between 1280 and 1546 of the S1 region (266 bp) (68). The primers used for PCR amplification of the S1 probe DNA were: 5'-S-5'-cccagcagagcagaggccactgaggagct-3' and 3'-S-5'-ggtcactgttgcccccctgcctgtcctggc-3'.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Ramos 2G6 B cell line is a model for IgM⁺ B cell responses to T cell signals

To study the relationship between IL-4- and CD40-mediated signaling and heavy chain class switching, we used a model system in which discrete aspects of B cell differentiation could be analyzed within a clonal population. The Ramos 2G6 B cell line retains many of the signaling pathways thought to mediate physiologic class switch events, as well as the genetic recombination/deletion mechanism involved in class switching in vivo. This line can up-regulate expression of I transcripts in response to IL-4, and undergoes limited switch recombination in response to IL-4 and CD40L (43). The Ramos 2G6 line is surface IgM⁺/IgD^- and secretes low levels of IgM Ab (52). Therefore, this B cell line appears to represent a B cell that has received an initial activation signal, but has not yet undergone switch recombination to express downstream isotypes.

Ramos 2G6 B cells remain IgM⁺ after growth in culture

To use the Ramos 2G6 cell line as a model system to study signals involved in switch recombination, we wished to confirm that this line was not undergoing isotype switching in culture. To establish the uniformity of the Ramos 2G6 B cell line with respect to Ig surface expression, we conducted two-color FACS analysis using either FITC-labeled -IgG or FITC-labeled -IgM in conjunction with PE-labeled -CD20. We found that the vast majority of the cells were positive for surface IgM (Fig. 1A) and negative for surface IgG expression (Fig. 1B). Additionally, supernatants from actively growing, unstimulated, Ramos 2G6 B cells lack IgG protein, as determined by ELISA (data not shown).

View larger version (20K): [in this window] [in a new window]   FIGURE 1. Expression of IgM or IgG on Ramos 2G6 B cells by FACS. Shown are FACS analyses of equal numbers of Ramos 2G6 cells stained with PE-labeled CD20 and either FITC-labeled IgM (A) or FITC-labeled IgG (B). The log10 fluorescence of each color is indicated on the appropriate axis of the two-color profiles.

Individual Ramos 2G6 subclones differ with respect to I transcription

An increase in steady state levels of germline transcripts by cytokines is tightly correlated with switch recombination to specific C_H loci (11, 12, 13, 14, 15, 16, 17, 18, 19). Because we wished to determine whether the Ramos 2G6 B cell line was actively producing I transcripts, and therefore primed to undergo switch recombination to a particular C_H region, we conducted S1 analysis on RNA isolated from unstimulated as well as IL-4- and T cell-stimulated Ramos 2G6 B cells. Previous analysis of Ramos 2G6 B cells by the S1 mapping technique revealed that germline I transcription was induced and augmented only after stimulation with either IL-4 or IL-4 together with an activated Jurkat cell line (D1.1), respectively. This previous study did not detect I transcription in unstimulated Ramos 2G6 B cells or in Ramos cells stimulated with D1.1 Jurkat T cells alone (43).

Using a probe that protected sequences in both the I and the C regions (Fig. 2B), we were able to confirm these previous observations. Initially we only detected germline (IC) transcripts after stimulation with IL-4 either alone or with the D1.1 cell line plus IL-4 (Fig. 2A, lanes 5 and 7). Mature (C) transcripts were detected only with IL-4 plus D1.1 cells (lane 7). However, in assaying different populations of Ramos 2G6 B cells after repeated passage in culture, we frequently detected a protected band that corresponded in size to the IC transcript (Fig. 2A, lanes 15, 17, and 18). This observation suggested that a basal level of I transcription was occurring in some cells before cytokine or T cell stimulation.

View larger version (39K): [in this window] [in a new window]   FIGURE 2. A, Effect of CD40L and IL-4 on the expression of I transcripts in Ramos 2G6 B cells. Ramos 2G6 B cells were incubated for 6 days under different conditions of coculture, and total RNA was isolated. Samples were analyzed by S1 nuclease protection for the expression of I and mature C transcripts (lanes 1–7). Shown are Ramos 2G6 cells cultured alone (lane 4); Ramos 2G6 cells cultured with 400 U/ml IL-4 (lane 5); Ramos 2G6 cells cultured CD40L+ D1.1 Jurkat T cells (lane 6); and Ramos 2G6 cells cultured with D1.1 T cells plus 400 U/ml IL-4 (lane 7). Also shown are pBR322/MspI markers (lane 1), undigested probe (lane 2), and a tRNA (yeast transfer RNA) control (lane 3). Different populations of unstimulated Ramos 2G6 cells were analyzed for basal levels of I expression (lanes 8–18). Total RNA from different isolates of Ramos 2G6 B cells was analyzed for expression of I and mature C transcripts. Protected bands corresponding to the IC and C transcripts are indicated by arrows on right. RNA protection assays were conducted with a uniformly labeled 440-bp IC probe schematically drawn in B. This probe is fully protected by the IC transcript to give a 440-bp band, and is partially protected by the mature C transcript to give a 300-bp band. This probe does not distinguish between transcripts from the different subclasses.

Subclass-specific germline transcript expression can be measured using RT-PCR

Because the S1 probe used to analyze Ramos 2G6 RNA after different conditions of coculture is cross-reactive with transcripts from all four C subclasses, we were unable to establish subclass-specific I expression (Fig. 2). To extend our earlier observations, we assessed the subclass-specific transcriptional response of I⁺ and I^- Ramos subclones in response to the CD40L⁺ Jurkat T cell line, D1.1. We used the RT-PCR protocol developed by Jumper et al. that is based on the use of 5' and 3' primers specific for homologous sequences within the four I and C genes (44). After amplification, I-specific PCR products are identified by hybridizing duplicate Southern blots with hinge region probes specific for each C subclass.

Because this PCR strategy depends on a particular hinge probe being specific for the indicated subclass, before performing experiments we wanted to confirm the specificity of each hinge region probe. We used isotype-specific primers to amplify and isolate the different I products from D1.1 plus IL-4-stimulated Ramos cells. After we confirmed the identity of each PCR product by sequencing, cloned fragments were transferred to duplicate nylon membranes, and probed with the different C hinge probes. We were particularly concerned with the possible cross-hybridization of the 1, 2, and 4 hinge probes because we were unable to unambiguously resolve the subclass-specific PCR products by size (1 = 428 bp, 2 = 416 bp, and 4 = 419 bp). However, as shown in Figure 3, we detected virtually no cross-hybridization among the 1, 2, and 4 probes. We did detect a low level of cross-hybridization between the 3 hinge probe and the amplified I1 fragment, but because these PCR products are readily separable by size (1 = 428 bp vs 3 = 569 bp), this level of cross-hybridization did not affect our ability to detect I1 and I3 transcripts.

View larger version (35K): [in this window] [in a new window]   FIGURE 3. Analysis of the specificity of the different C hinge oligo probes. PCR products representing I transcripts from the four C subclasses were cloned into pGEM-T plasmid, and each product identity was verified by sequencing. I PCR products (designated at top of figure) were digested and released as fragments, and digested products were run on duplicate Southern blots. Each blot was probed with the specific [-32P]-kinased hinge region oligo indicated on the vertical axis.

Distinct subclones of Ramos 2G6 are transcribing basal levels of I1, I2, and I3, but not I4 transcripts

To further characterize the basal transcription in an I⁺ population, one line (C2) was subcloned and independent subclones were analyzed by RT-PCR for subclass-specific I expression. We found that all but one subclone expressed I1 transcripts. Subsets of I1⁺ subclones also expressed either 1) I2 or I3 transcripts or 2) both I2 and I3 transcripts. We did not detect any I4 expression in any of the subclones tested (Fig. 4). Further analysis of the I^- subclone after growth in culture revealed that it eventually began to express detectable I transcripts. This suggested that the Ramos 2G6 subclone was not stable with respect to I transcription, but was becoming I⁺ upon growth in culture.

View larger version (54K): [in this window] [in a new window]   FIGURE 4. Subclass-specific expression of secondary subclones of an I+ Ramos 2G6 cell line, C2. The C2 line was subcloned by limiting dilution, and multiple subclones were assayed for basal expression of I transcripts. RNA was isolated from the different subclones, and RT-PCR was conducted with primers specific for the I transcripts. Duplicate blots were hybridized with hinge region probes that were kinased to a similar sp. act. (subclass indicated on left side of figure). Exposure of all four blots was for an equal length of time.

I1⁺ subclones show a unique pattern of regulation in response to IL-4 and T cell contact

To assess whether an I⁺ subclone (D9) up-regulates germline transcripts in response to T cell contact, we established cocultures with Jurkat T cell lines that either express (D1.1) or do not express (B2.7) the CD40L (57, 58). Using the anti-CD40L mAb, 5c8, we confirmed the CD40L expression pattern of the B2.7 and D1.1 Jurkat cell lines (Fig. 5A, top and middle panels). After 6 days, RNA was isolated and analyzed by RT-PCR and hybridization with hinge region-specific probes. This analysis revealed that expression of I1, I2, and I3 transcripts was markedly increased in Ramos D9 cultures incubated with IL-4 (Fig. 5B, lane 2). In addition, I transcription from these three subclasses was up-regulated in response to the CD40L⁺ Jurkat line, D1.1 (lane 5). With D1.1 T cells plus IL-4, we observed an augmentation of I transcription from the 1, 2, and 3 loci over what was seen with IL-4 alone (Fig. 5B, compare lane 6 with lane 2). We also observed after a long exposure a low level increase in I4 transcripts with IL-4 and/or CD40L (data not shown).

View larger version (39K): [in this window] [in a new window]   FIGURE 5. Analysis of germline expression in an I+ subclone (D9) in response to different conditions of coculture. A, Shown are fluorescence histogram (FACS) analyses of the CD40L- B2.7 and the CD40L+ D1.1 Jurkat T cell lines, as well as the CD40L+ 293/CD40L-transfected cell line. The y-axis represents number of cells, and the x-axis represents relative fluorescence intensity using the 5c8 (anti-CD40L) mAb (broken line). The solid line in all graphs represents cells stained with the second Ab alone (region 1). The horizontal bar in each histogram identifies 5c8-positive cells (region 2) that are designated as a percentage of total cells analyzed (regions 1 plus 2). B, RNA from 6-day-old cultures was isolated and used in RT-PCR assays containing sequence primers specific for I and VDJ-Cµ transcripts. Amplified products were analyzed by Southern blotting and hybridization with hinge region-specific probes for the different subclasses (probes are indicated on left). The hinge probes were kinased to approximately the same sp. act. The 1, 2, and 3 blots were exposed overnight, and in this experiment the 4 blot was exposed for 6 days. Shown are D9 B cells incubated alone (lane 1); with IL-4 (lane 2); with mitomycin-treated B2.7 Jurkat T cells (CD40L-) (lane 3); with B2.7 T cells in addition to IL-4 (lane 4); with mitomycin-treated D1.1 Jurkat T cells alone (lane 5); and with D1.1 T cells plus IL-4 (lane 6). Cµ expression, shown in the lower panel, is a control for the fidelity of the RT-PCR reaction in each individual sample. C, Shown is the result of a coculturing experiment using Ramos-D9 B cells and 293 cells or 293/CD40L cells, in the presence or absence of IL-4. RNA was isolated from the cocultures and assayed for I1 expression. The different coculturing conditions of the Ramos 2G6 D9 subclone were as follows: Ramos 2G6 D9 cells alone (lane 1); with 400 U/ml IL-4 (lane 2); with mitomycin-treated 293/CD40L cells (lane 3); with mitomycin-treated 293 cells (lane 4); with mitomycin-treated 293/CD40L cells plus IL-4 (lane 5); and with mitomycin-treated 293 cells plus IL-4 (lane 6). Experiments shown were repeated twice with independent I+ subclones.

Because it has been shown previously that specific B cell subsets and B cell lines can express a low level of CD40L upon activation (69, 70, 71), we wanted to rule out the possibility that CD40L expression by the Ramos 2G6 line was contributing to the response we observed. Using an Ab against CD40L on Ramos 2G6 cells either before or after IL-4 stimulation, we were unable to detect by FACS any surface expression of CD40L (data not shown).

CD40L stimulation alone is a poor stimulator of I transcription in the I^- subclone

We next assessed the transcriptional activation of the loci in a Ramos 2G6 subclone that showed no basal level of I transcription. A representative experiment with one I^- subclone, designated C8, is shown in Figure 6.

View larger version (61K): [in this window] [in a new window]   FIGURE 6. Effect of IL-4- and CD40L-expressing and nonexpressing Jurkat T cell lines on the expression of I transcripts in an I- Ramos 2G6 subclone. A, The I- subclone was incubated for 6 days under different conditions and assayed for I expression by RT-PCR. The conditions of incubation in lanes 1 to 6 are as follows: Ramos 2G6 B cells alone (lane 1); Ramos 2G6 B cells incubated with 400 U/ml IL-4 (lane 2); Ramos 2G6 B cells incubated with 5 x 105 D1.1 cells (lane 3); Ramos 2G6 B cells incubated with 5 x 105 B2.7 cells (lane 4); Ramos 2G6 B cells incubated with IL-4 + D1.1 cells (lane 5); and Ramos 2G6 B cells incubated with IL-4 + B2.7 cells (lane 6). All four subclass blots were exposed for 24 h. To control for RNA integrity and the efficacy of the PCR reaction, primers specific for regions in the Cµ message were included in each PCR reaction. To identify VDJ-Cµ expression, one I blot was reprobed with an end-labeled JH-specific probe (bottom panel). B, A 1-wk exposure of the I1 blot shown in Figure 7A to show expression of I1 transcripts by D1.1 stimulation alone. C, Ramos 2G6 I- B cells were incubated with mitomycin-treated 293 or 293/CD40L cells. RNA was isolated from the cocultures and assayed for I1 expression. The lanes represent the following conditions: Ramos 2G6 I- subclone (C8) alone (lane 1); C8 cells incubated with 400 U/ml IL-4 (lane 2); with 5 x 105 293 cells (lane 3); with 5 x 105 293/CD40L cells (lane 4); with 5 x 105 293 cells plus 400 U/ml IL-4 (lane 5); and with 5 x 105 293/CD40L cells plus IL-4 (lane 6). PCRs were conducted with primers specific for both IC and Cµ cDNAs. Cµ PCR products are shown in the bottom panel. Data are representative of one of three separate experiments conducted.

Together these experiments reveal that Ramos 2G6 B cells with an I^- phenotype generally do not up-regulate germline transcripts in response to CD40L⁺ D1.1 T cells alone. This result is in direct contrast to what we observed in the I⁺ subclones incubated under the same conditions (Fig. 5B, lane 5). However, under the same conditions and in the presence of IL-4, both the I^- and I⁺ populations augment the expression of I transcripts to a level that is significantly greater than what is observed with IL-4 alone.

To establish that our results were the consequence of CD40 signaling and not a result of other lymphoid-specific molecule(s), we again repeated our experiments using 293/CD40L cells. In the presence of 293/CD40L cells plus IL-4, we observed increased induction of I1 transcripts relative to stimulation with 293 plus IL-4 (Fig. 6C, compare lanes 5 and 6). We also observed a modest induction of transcription in response to CD40L alone (lane 4). Since we could also detect a slight induction of I1 transcripts in the I^- population by D1.1 T cells after a long exposure (Fig. 6B), one possibility is that there is a small number of I⁺ cells that can up-regulate germline transcripts in response to CD40L. To test this possibility, we subcloned an I^- subclone and analyzed 10 secondary subclones for basal I1 expression. We identified two subclones that were positive for low level basal I transcription (data not shown). Therefore, the low level of I1 germline transcription that is seen in response to CD40 signaling is consistent with the presence of a small number of I-expressing cells present within the I^- population.

Quantitative measurements of I1 transcription after incubation of I^- and I⁺ Ramos 2G6 B cells with IL-4 and Jurkat T cells

To quantitate the up-regulation of I1 transcripts under different conditions of stimulation, we conducted semiquantitative RT-PCR on RNA isolated from cocultures established with I^- (C8) and I⁺ (D9₄) Ramos B cells. To control for B cell-specific RNA in the PCR reactions, we coamplified a fragment that corresponded to sequences in the VDJ-Cµ mRNA. The extent of augmentation was determined by quantifying both signals within a linear range of amplification and dividing the I1 signal by the Cµ signal (Fig. 7).

View larger version (48K): [in this window] [in a new window]   FIGURE 7. Semiquantitative PCR analysis of RNAs isolated from an I- subclone (C8) and an I+ subclone (D94) after stimulation with T cells and IL-4. To quantitate the induction of I1 transcripts in the C8 and D94 subclones after stimulation, RNA was extracted 6 days after coculture. RNA was reverse transcribed, amplified by PCR, and analyzed for I1 expression by hybridization to the 1 hinge probe. Shown are the different cDNA samples representing an I- and an I+ phenotype diluted over a 2-log range and then amplified by PCR. Blots were hybridized overnight with the C1 hinge probe. In the same reaction, primers to amplify the Cµ message were added to quantify and standardize the amount of B cell-specific RNA in each reaction. Shown are PCR products from the C8 line (A) or the D94 line (B). PCR reactions were conducted with 5 µl (A, lanes 1, 4, and 7; B, lanes 1, 4, 7, 10, and 13), 0.5 µl (A, lanes 2, 5, and 8; B, lanes 2, 5, 8, 11, and 14), and 0.05 µl (A, lanes 3, 6, and 9; B, lanes 3, 6, 9, 12, and 15) of cDNA template. Band intensities were determined by quantitation using a phosphor imager. To quantitate the two signals, the band intensities were plotted, and values that fell in a linear range of input cDNA to product were used to determine the I expression/Cµ ratio. In all cases, the first or second dilution was used to quantitate the I/Cµ signal.

Analysis of I expression in the I⁺ subclone D9₄ showed that IL-4 increased the basal I1 expression approximately fivefold over the signal obtained with unstimulated cells (Fig. 7B). After incubating the B cells with CD40L⁺ D1.1 T cells, we observed a similar induction of I1 transcripts over the unstimulated D9₄ B cells (approximately sixfold). In contrast, there was no discernible increase in I1 expression after incubating the D9₄ B cells with the B2.7 T cells (see lanes 7–9). In confirmation of others’ findings (43, 44, 47, 49, 55, 56), incubation of both the C8 and D9₄ B cell lines with IL-4 plus D1.1 T cells produced an increase in I1 expression that was greater than what was measured with either IL-4 or CD40L⁺ D1.1 T cells alone.

Switch recombination in an I⁺ subclone requires both IL-4 and CD40L

In an effort to relate the expression of I transcripts in the Ramos 2G6 line to isotype class switching, we assayed the expression of mature C transcripts in response to different conditions of coculture. In Figure 8A, a subclone, RA3, that displays a relatively high basal level of I1 expression and a very low level of I2 expression was analyzed for the induction of germline transcription in response to different conditions of coculture. Upon culturing this line with 293/CD40L cells, we observed the up-regulation of I1, and to a much lesser degree, I2 and I4 transcription compared with what is observed with 293 cells alone (compare lanes 4 and 3, respectively). After longer exposures, we could also observe the induction of I3 transcripts (data not shown). We repeated these experiments using semiquantitative PCR and assessed the expression of all four subclasses. We analyzed the expression of I and Cµ transcripts at a time when these signals were increasing linearly with respect to input cDNA. Upon quantification of the different I signals, it was evident that there was a clear synergistic effect of CD40L and IL-4 on the different loci. However, the magnitude of expression was much greater at the 1 locus compared with the other three loci (Table I).

View larger version (53K): [in this window] [in a new window]   FIGURE 8. Ramos 2G6 B cells express mature C1 transcripts only under conditions of IL-4 and CD40L stimulation. To establish the conditions of stimulation that would induce mature C transcripts in an I+ (RA3) population and the subclass-specific expression of these transcripts, RA3 cells were incubated under different coculture conditions of stimulation. Shown in A is the amplification of cDNA with primers specific for I sequences. In B is the result of PCR amplification of the same cDNAs using primers specific for mature C transcript. The different conditions of coculture are as follows: alone (lane 1); with IL-4 (lane 2); with mitomycin-treated 293 cells (lanes 3); with mitomycin-treated 293 cells plus IL-4 (lanes 4); with mitomycin-treated 293/CD40L cells (lanes 5); and with 293/CD40L cells plus IL-4 (lanes 6). All hinge region probes were labeled to the same approximate sp. act. The exposure of all blots within an experiment was for the same length of time. To control for RNA integrity and the efficacy of the PCR reaction, primers specific for regions in the Cµ message were included in each PCR reaction. One I blot was reprobed with the 5'-JH oligo to identify the Cµ-specific band.

View this table: [in this window] [in a new window]   Table I. Quantitation of subclass I transcripts in the RA3 subclone after stimulation with IL-4, CD40L, and CD40L plus IL-4

To confirm that we were inducing de novo switch recombination and not just selecting previously switched cells, we analyzed DNA from cocultures of Ramos 2G6 using an assay that detects the reciprocal recombination products of switch recombination events (46, 66, 68, 72). As shown in Figure 9, we only detected switch circle products in samples of Ramos 2G6 B cells that had been cocultured with 293/CD40L cells plus IL-4. In addition, the presence of multiple discrete bands is consistent with several Ramos B cells, with distinct Sµ-S breakpoints, having undergone switch recombination as a consequence of CD40L plus IL-4 signaling.

View larger version (26K): [in this window] [in a new window]   FIGURE 9. Detection of PCR-amplified products containing S sequences from switch circle DNA. PCR products amplified with primers specific for reciprocal switch products were transferred to nylon and probed with an S-specific probe. Shown is a probed blot of PCR products amplified from DNA isolated from Ramos 2G6 B cells cultured either alone (lane 2) with 293/CD40L (lane 3), or with 293/CD40L plus IL-4 (lane 4). Molecular weight markers ( HindIII/EcoRI) are shown in lane 1.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Using this model system, we found that D1.1 T cells efficiently induce or up-regulate I transcription in Ramos 2G6 subclones that display a basal level of I transcription. We found in I⁺ subclones that I expression was up-regulated at multiple loci, irrespective of the transcriptional state of a particular locus before coculturing. For example, in a subclone that only expressed detectable basal levels of I1 transcripts, after incubation with D1.1 T cells alone, we observed the induction of I2 and I3 as well as I1 transcripts. In contrast, we found that the CD40L failed to induce germline transcription at the 2, 3, and 4 loci, and induced very weak transcription at the 1 locus in the Ramos 2G6 I^- subclones.

This finding was reinforced when we analyzed the induction of I transcripts using semiquantitative PCR. Whereas the induction of I1 transcripts in the transcriptionally positive subclone was approximately the same with either IL-4 or CD40L, the I^- subclone was induced to a much greater extent with IL-4 compared with the minimal induction seen with CD40L contact alone.

Together these results suggest that only after the Ramos B cells have undergone a change that is reflected in the induction of basal C_H transcription can I transcripts be effectively induced by CD40 signaling. To extend this model to normal B cell responses, the I⁺ phenotype appears to represent a stage of B cell activation in which there is a general increase in B cell responsiveness to CD40 signaling.

This model could explain differences obtained in analogous studies in humans and mouse B cells. Kitani and Strober studying I expression in IgG^- high density resting B cells found that I1 and I3 expression generally required a proliferative stimulus, whereas I2 and I4 transcripts could be induced with cytokines alone (IFN- and IL-4, respectively) (73). However, Fujieda et al. found that in IgD⁺ tonsillar B cells, I1, I3, and I4, but not I2 transcripts were induced with IL-4 alone (46). This same study found that CD40 triggering alone failed to induce the expression of any class of I transcripts. However, Jumper et al. (44) and Warren and Berton (49) found that multiple classes of germline transcripts were up-regulated in response to CD40 signaling. One possible explanation for these contrasting results is that the different methods used to isolate B cells resulted in populations or subpopulations at different stages of B cell activation also differed in their ability to up-regulate germline transcripts in response to CD40 triggering.

Other evidence also supports the concept that up-regulation of germline transcripts by CD40 signaling may be contingent upon the activation state of the B cell. Berberich et al. found that transformed B cell lines, compared with human tonsillar B cells, differentially activated a number of kinases in response to signaling through CD40 (40). The capacity to activate specific signaling pathways correlated with a specific maturation stage of the responding B cells.

The up-regulation of I transcripts by stimulation with either IL-4 plus CD40L or CD40L alone suggests that CD40-responsive elements are associated with different C genes. Such response elements recently have been identified in the promoter region of the human I gene (51) and the murine I1 gene (41). It is possible that signaling through CD40 induces the activation of a unique set of transcription factors distinct from those induced with IL-4. Alternatively, because IL-4 and CD40 signaling are known to up-regulate a number of common transcription factors, including nuclear factor-B, AP-1, and nuclear factor of activated T cells (NF-AT) (39, 74, 75, 76), signals provided by coactivation could increase the number of occupied binding sites at each I promoter, resulting in increased transcription.

Switch recombination, as measured by the appearance of both mature C transcripts and reciprocal products, was observed only after Ramos 2G6 B cells were stimulated with both IL-4 and CD40L. This result supports others’ findings that both signals are necessary for switch recombination (43, 44, 47, 49, 55, 56). However, contact through CD40, which results in increased germline transcription, is not sufficient to obtain switch recombination in our model system. This is evident in our I⁺ subclone, RA3. This subclone expresses an elevated basal level of I1 transcription and clearly up-regulates the expression of I1 transcripts in response to CD40L, but does not undergo switch recombination when stimulated through CD40 alone. This result contrasts with the findings of Malisan et al., who detected a low level of switch recombination in IgD⁺ B cells cocultured with CD40L-expressing L cells. Whether these rare events represented a small number of B cells that were already cytokine and/or Ag activated in the population is unknown. It is clear, however, that the majority of B cells undergoing switch recombination did require both IL-10 and CD40L (66).

We found in the RA3 line that transcription was up-regulated at the I1 locus in response to CD40L plus IL-4 approximately 14-fold over the basal level of transcription. This level of induction was not approached at the 1 locus as a consequence of other culturing conditions or at the other loci in response to any condition of activation (Table I). Since we only saw measurable levels of VDJ-C1 transcripts, it is possible, as suggested by Snapper et al. (77), that recombination occurs only after a critical threshold of germline transcription is reached. Alternatively, IL-4 and CD40L signaling may induce factors actually involved in the switching mechanism.

In conclusion, we find a differential effect of CD40 signaling on the induction of I transcripts in phenotypically distinct subclones of Ramos 2G6. This difference in signaling may ultimately reflect chromatin changes at the C_H locus that facilitate or prevent the binding of CD40-induced transcription factors. Chromatin changes have been seen previously at the C_H locus in response to cytokine signaling (8, 9). We are currently examining whether analogous changes occur in our Ramos clones as they move from an I^- to an I⁺ phenotype.

	Acknowledgments

 
We gratefully acknowledge Dr. Seth Lederman for providing the Jurkat T cell lines and the 5c8 mAb, Dr. Paul Rothman for the IC3 S₁ probe, and Dr. Frederick Mills for the Sµ-S1 control plasmid and help with technical aspects of the switch product PCR protocol. We also thank Dr. Suzanne Li for critically reading the manuscript, and Ameesha Bhushan and Dr. Suzanne Li for many useful discussions.

	Footnotes

 
¹ This work was supported in part by National Institutes of Health Grant R2AI37081 to L.R.C., and by a grant from Charles and Johanna Busch Memorial Fund, Rutgers, The State University of New Jersey.

² Address correspondence and reprint requests to Dr. Lori R. Covey, Nelson Biologic Laboratories, Rutgers, The State University of New Jersey, Piscataway, NJ 08855. E-mail address:

³ Abbreviations used in this paper: TD, T cell-dependent; CD40L, CD40 ligand; PE, phycoerythrin; RT-PCR, reverse-transcriptase polymerase chain reaction; S, switch.

Received for publication January 17, 1997. Accepted for publication September 25, 1997.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Banchereau, J., F. Bazan, D. Blanchard, F. Briere, J. P. Galizzi, C. v. Kooten, Y. J. Liu, F. Rousset, S. Saeland. 1994. The CD40 antigen and its ligand. Annu. Rev. Immunol. 12:881.[Medline]
2. Armitage, R. J., W. C. Fanslow, L. Strockbine, T. A. Sato, K. N. Clifford, B. M. Macduff, D. M. Anderson, S. D. Gimpel, T. Davis-Smith, C. R. Maliszewski, E. A. Clark, C. A. Smith, K. H. Grabstein, D. Cosman, M. K. Spriggs. 1992. Moelcular and biological characterization of a murine ligand for CD40. Nature 357:8082.
3. Fanslow, W. C., K. N. Clifford, M. Seaman, M. R. Alderson, M. K. Spriggs, R. J. Armitage, F. Ramsdell. 1994. Recombinant CD40 ligand exerts potent biologic effects on T cells. J. Immunol. 152:4262.[Abstract]
4. Fanslow, W. C., D. M. Anderson, K. H. Grabstein, E. A. Clark, D. Cosman, R. J. Armitage. 1992. Soluble forms of CD40 inhibit biologic responses of human B cells. J. Immunol. 149:655.[Abstract]
5. Hollenbaugh, D., L. S. Grosmaire, C. D. Kullas, N. J. Chalupny, S. Braesch-Andersen, R. J. Noelle, I. Stamenkovic, J. A. Ledbetter, A. Aruffo. 1992. The human T cell antigen gp39, a member of the TNF gene family, is a ligand for the CD40 receptor: expression of a soluble form of gp39 with B cell co-stimulatory activity. EMBO J. 11:4313.[Medline]
6. Spriggs, M. K., R. J. Armitage, L. Strockbine, K. N. Clifford, B. M. Macduff, T. A. Sato, C. R. Maliszewski, W. C. Fanslow. 1992. Recombinant human CD40 ligand stimulates B cell proliferation and immunoglobulin E secretion. J. Exp. Med. 176:1543.[Abstract/Free Full Text]
7. Coffman, R. L., H. F. J. Savelkoul, D. A. Lebman. 1989. Cytokine regulation of immunoglobulin isotype switching and expression. Semin. Immunol. 1:55.
8. Schmitz, J., A. Radbruch. 1989. An interleukin-4-induced DNase I hypersensitive site indicates opening of the 1 switch region prior to switch recombination. Int. Immunol. 1:570.[Abstract/Free Full Text]
9. Berton, M. T., E. S. Vitetta. 1990. Interleukin 4 induces changes in the chromatin structure of the 1 switch region in resting B cells before switch recombination. J. Exp. Med. 172:375.[Abstract/Free Full Text]
10. 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]
11. 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]
12. Lutzker, S., P. Rothman, R. Pollock, R. Coffman, F. W. Alt. 1988. Mitogen and IL-4 regulated expression of germline Ig 2b transcripts evidence for directed heavy chain class switching. Cell 53:177.[Medline]
13. Rothman, P., S. Lutzker, B. Gorham, V. Stewart, R. Coffman, F. W. Alt. 1990. Structure and expression of germline immunoglobulin 3 heavy chain gene transcripts: implications for mitogen and lymphokine directed class switching. Int. Immunol. 2:621.[Abstract/Free Full Text]
14. Xu, M., J. Stavenezer. 1990. Structure and expression of germ line immunoglobulin heavy-chain gamma 1 transcripts in interleukin 4 treated mouse spleen cells. Dev. Immunol. 1:11.[Medline]
15. Rothman, P., Y. Y. Chen, S. Lutzker, S. C. Li, V. Stewart, R. Coffman, F. W. Alt. 1990. Structure and expression of germ line immunoglobulin heavy-chain epsilon transcripts: interleukin-4 plus lipopolysaccharide-directed switching to C epsilon. Mol. Cell. Biol. 10:1672.[Abstract/Free Full Text]
16. 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]
17. Xu, L., B. Gorham, S. C. Li, A. Bottaro, F. W. Alt, P. Rothman. 1993. Replacement of germ-line promoter by gene targeting alters control of immunoglobulin heavy chain class switching. Proc. Natl. Acad. Sci. USA 90:3705.[Abstract/Free Full Text]
18. Zhang, J., A. Bottaro, S. Li, V. Stewart, F. W. Alt. 1993. A selective defect in IgG2b switching as a result of targeted mutation of the I2b promoter and exon. EMBO J. 12:3529.[Medline]
19. Jung, S., K. Rajewsky,