Blood Subsets from Human Tonsil and Peripheral RAG1 and RAG2 ...

RAG1 and RAG2 Expression by B Cell Subsets from Human Tonsil and Peripheral Blood1 Hermann J. Girschick, Amrie C. Grammer,2 Toshihiro Nanki,2 Marlyn Mayo, and Peter E. Lipsky2,3 It has been suggested that B cells acquire the capacity for secondary V(D)J recombination during germinal center (GC) reactions. The nature of these B cells remains controversial. Subsets of tonsil and blood B cells and also individual B cells were examined for the expression of recombination-activating gene (RAG) mRNA. Semiquantitative analysis indicated that RAG1 mRNA was present in all tonsil B cell subsets, with the largest amount found in naive B cells. RAG2 mRNA was only found in tonsil naive B cells, centrocytes, and to a lesser extent in centroblasts. Neither RAG1 nor RAG2 mRNA was routinely found in normal peripheral blood B cells. In individual tonsil B cells, RAG1 and RAG2 mRNAs were found in 18% of naive B cells, 22% of GC founder cells, 0% of centroblasts, 13% of centrocytes, and 9% of memory B cells. Individual naive tonsil B cells containing both RAG1 and RAG2 mRNA were activated (CD69ⴙ). In normal peripheral blood ⬃5% of B cells expressed both RAG1 and RAG2. These cells were uniformly postswitch memory B cells as documented by the coexpression of IgG mRNA. These results indicate that coordinate RAG expression is not found in normal peripheral naive B cells but is up-regulated in naive B cells which are activated in the tonsil. With the exception of centroblasts, RAG1 and RAG2 expression can be found in all components of the GC, including postswitch memory B cells, some of which may circulate in the blood of normal subjects. The Journal of Immunology, 2001, 166: 377–386.

T

he V(D)J rearrangement of immune receptor genes is a feature of developing lymphocytes in the bone marrow (1) and depends upon recombination-activating gene (RAG)4 enzymes (2). The proteins encoded by RAG1 and RAG2 are exclusively responsible for the sequence-specific recognition of conserved recombination signal sequences flanking the rearranging gene elements and cleavage of DNA at these regions precisely (3–7). RAG enzymes are expressed at high levels during ontogeny, but their expression diminishes in immature B cells and is usually absent in recirculating mature naive B cells (8, 9). Mature naive B cells can be activated by Ag in the T cell-rich zones of secondary lymphoid organs, initiating the formation of germinal centers (GC) in the primary B cell follicles (10 –13) that promote clonal expansion (14), somatic hypermutation (15–17), and switch recombination of Ag-specific B cells (18 –20). B cells involved in the GC reaction appear to regain phenotypic and molecular traits of immature cells (21), including the re-expression of RAG-enzymes, potentially permitting secondary V(D)J rearrangement of Department of Internal Medicine, Harold C. Simmons Arthritis Research Center, University of Texas Southwestern Medical Center, Dallas, TX 75235 Received for publication February 2, 2000. Accepted for publication October 9, 2000. 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 National Institutes of Health Grant AI 31229. A.C.G. was supported by National Institutes of Health Postdoctoral Training Grant AR-18550. H.J.G. is a recipient of a Deutsche Forschungsgemeinschaft Grant Gi 295/1-1. 2 Current address: National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, 9000 Rockville Pike, Bethesda, MD 20892-1820. 3 Address correspondence and reprint requests to Dr. Peter E. Lipsky, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, 9000 Rockville Pike, Bethesda, MD 20892-1820. E-mail address: [email protected]

Ig genes (22–31). In addition to re-expression of RAG enzymes, expression of TdT (32) and of surrogate (␺) light chain has been demonstrated in human GC reactions (30), although TdT does not appear to be expressed in murine GC reactions (23). It has been proposed that secondary V(D)J recombination in secondary lymphoid organs might be a mechanism to rescue B cells whose Ag receptor avidity has been decreased as a result of somatic hypermutation (25). Recent evidence using transgenic mice expressing RAG promoter constructs, however, has questioned whether transcription of RAG genes is up-regulated in secondary lymphoid organs. In this regard, transcription of RAG genes was limited to B cell precursor cells and transitional cells that had left the bone marrow and were completing maturation in the spleen (33, 34). We had previously reported evidence that patients with a systemic autoimmune disease, systemic lupus erythematosus, manifested evidence of secondary rearrangement of V␬ genes in the periphery (35, 36). Therefore, it was important to determine whether B cells re-expressed RAG enzymes in secondary lymphoid organs during immune responses. RAG expression was assessed by using a combination of multiparameter flow cytometric cell sorting and RT-PCR analysis of tonsil and peripheral blood B cell mRNA content at both the subset and individual cell level. RAG1 and RAG2 mRNA expression by activated naive B cells in the tonsil as well as by B cells at subsequent stages of maturation in the GC was documented. Moreover, postswitch IgG expressing activated memory B cells in both blood and tonsil expressed RAG1 and RAG2 mRNA. These results imply that many B cells generated during GC reactions may have the molecular machinery to accomplish secondary V(D)J rearrangement.

Materials and Methods Preparation of B cells from peripheral blood

4

Abbreviations used in this paper: RAG, recombination-activating gene; GC, germinal center; OGG, 8-oxoguanine DNA glycosylase; RT, reverse transcriptase; s, surface; G6PD, glucose-6-phosphate dehydrogenase. Copyright © 2001 by The American Association of Immunologists

PBMC were separated by Ficoll-Hypaque density gradient centrifugation from heparinized peripheral blood of 14 healthy adult Caucasian donors. B 0022-1767/01/$02.00

378

RAG EXPRESSION IN HUMAN TONSIL AND PERIPHERAL BLOOD B CELLS

cells were purified from PBMCs by negative selection using magnetic separation according to the manufacturer’s instructions (Stemcell Technologies, Vancouver, British Columbia, Canada). Purity of the B cell enrichment was determined by indirect immunofluorescence staining using a PEconjugated anti-human CD19 mAb (PharMingen, San Diego, CA) and analysis with a FACScan flow cytometer (Becton Dickinson, San Jose, CA). B cells in the 14 different samples were 93–96% CD19 positive.

Preparation of B cells from tonsil tissue and of mononuclear cells from thymus Human tonsils from six healthy children were obtained after routine tonsillectomy. Suspensions of tonsilar mononuclear cells were prepared by collagenase digestion (Worthington Biochemical, Lakewood, NJ) of the tissue for 30 min, followed by Ficoll-Hypaque density gradient centrifugation, as described previously (32). Human thymus tissue from one otherwise healthy child was obtained during heart surgery for congenital heart disease. Suspensions of thymus mononuclear cells were prepared as described above.

Cell sorting Tonsil B cells subsets from six subjects were defined by the surface expression of CD19, IgD, and CD38 and were isolated with a FACStarPlus or FACSVantage flow cytometer (Becton Dickinson) as described previously (32) (see Fig. 1). Indirect immunofluorescence staining was conducted using a biotinylated anti-human CD19 mAb (Coulter, Miami, FL) followed by streptavidin-RED670 (Life Technologies, Gaithersburg, MD), a PEconjugated anti-human CD38 mAb (Becton Dickinson), and a FITC-labeled goat anti-human IgD polyclonal Ab (Caltag, San Francisco, CA). Postsort analysis of isolated populations revealed a purity of ⬎90%.

Single-cell sorting For the deposition of individual cells into microtiter wells, tonsilar mononuclear cells were stained as described above. Single cells were sorted into 96-well PCR plates (Robbins Scientific, Sunnyvale, CA) using a FACStarPlus flow cytometer (Becton Dickinson) outfitted with a single-cell deposition unit, as described previously (37). For the sorting of tonsil naive and memory B cells into CD69-positive and -negative cells (38), four-color immunofluorescence staining was performed using a biotinylated anti-human CD19 mAb (Coulter) followed by streptavidin Per CP (Becton Dickinson), a PE-conjugated anti-human CD38 mAb (Becton Dickinson), a FITC-labeled goat anti-human IgD polyclonal Ab (Caltag), and a tricolor-conjugated anti-human CD69 mAb (Caltag). Single cells were sorted with a FACSVantage flow cytometry system (Becton Dickinson). In some experiments, cells were sorted in bulk and then deposited in the wells manually by limiting dilution.

Preparation of RNA and cDNA from sorted bulk populations Total RNA was extracted using the RNAeasy RNA isolation kit (Qiagen, Chatsworth, CA). The cell numbers subjected to RNA preparation were as follows: 0.5 ⫻ 106 FACS-sorted B cells of each tonsil subset and 2.0 ⫻ 106 negatively selected B cells from normal peripheral blood of each donor. Contaminating genomic DNA was removed using RQ1 RNase-free DNase according to the manufacturer’s instructions (Promega, Madison, WI). For conversion of mRNA into cDNA, Superscript II RNase H-reverse transcriptase (RT; Life Technologies) was used according to the manufacturer’s instructions. From each RNA, one sample was prepared by omitting the RT and used as control.

Preparation of RNA and cDNA from sorted single cells A total of 5 ␮l of lysis solution (1 ␮l of 5⫻ first-strand buffer B (Life Technologies), 0.01 M DTT (Life Technologies), 1% Nonidet P-40 (Sigma, St. Louis, MO), 5 U of recombinant RNasin ribonuclease inhibitor (Promega), 800 ␮M each dATP, dCTP, dGTP, and dTTP (Sigma), 0.05 ␮g of oligo(dT)12–18 (Pharmacia Biotech, Piscataway, NJ)) was added into each well of the PCR plate before sorting individual cells into the wells. The conversion of mRNA from a single cell was conducted with Superscript II RNase H-RT (Life Technologies) as described previously (1). In short, after annealing for 1 min at 65°C, followed by a 2-min incubation at 20°C, 3 ␮l of RT reaction mix (0.6 ␮l of 5⫻ first-strand buffer B, 0.01 M DTT, 10 U of RT) was added. Reverse transcription was conducted for 50 min at 42°C, followed by heat inactivation at 65°C for 10 min. For removal of the remaining mRNA from the newly synthesized cDNA, RNase H (0.08 U/reaction; Amersham Life Science, Cleveland, OH) was used.

PCR amplification The relative amount of RAG1-, RAG2-, TdT-, 8-oxoguanine DNA glycosylase (OGG)-, VpreB-, 14.1- encoding cDNA in the different samples was determined by amplification of glucose-6-phosphate dehydrogenase (G6PD) or ␤-actin adjusted samples using the PCR Southern blot technique, as described below. The primer pairs used are listed in Table I. In the first round of PCR, 30 (OGG, VpreB, 14.1, IgD, IgM, IgG) or 38 cycles (RAG1 and RAG2, TdT) of 30 s at 94°C, 1 min at 60°C, and 2 min at 72°C with a final 5-min extension at 72°C were conducted. A second nested step of PCR was performed only in the analysis of cDNA generated from single cells. Five microliters of the first PCR product was subjected to nested PCR amplification using primers as indicated in Table I. For analysis of human RAG2 mRNA, two different 5⬘ primers for the alternative exon 1a and exon 1b were used. The RAG2 exon 1a primers were intron spanning, whereas the RAG1 and RAG2 exon 1b forward primers spanned an intron. These primers were chosen because completely intron spanning primers for RAG2 exon 1b and RAG1 did not yield reproducible results with singlecell RT-PCR. Control samples of each preparation carried out without RT were run in parallel and were negative for the presence of contaminating DNA (data not shown). In addition, PCR without adding cDNA were run in parallel and did not yield a product.

Detection of amplified cDNA by PCR Southern blot analysis For quantification of cDNA in the different samples, PCR amplification for the housekeeping genes, G6PD or ␤-actin, was done in triplicate at 20, 25, 30, and 35 PCR cycles each. RT-PCR products were analyzed on a 1.8% agarose gel and transferred to a nylon membrane by alkaline vacuum transfer (Bio-Rad, Hercules, CA). PCR Southern blots were incubated in hybridization fluid containing ␥-32P-labeled probes for G6PD or ␤-actin PCR products, as listed in Table II. The amount of hybridized probe for the housekeeping gene was quantified by phosphor imager. The linear range of PCR was determined by plotting the amount of phosphor imager counts (above background) of the hybridized probe as a function of the number of cycles run. The mean counts of these triplicates in the linear range of PCR amplification were calculated and were used to calculate an amount of cDNA (in ␮l) of each preparation that contained the same relative content of housekeeping gene cDNA. This calculated volume of cDNA was subjected to further PCR amplification for specific cDNAs (39). The amplified PCR products were again analyzed by agarose gel electrophoresis and transferred to a nylon membrane and blotted. The number of PCR cycles in which the PCR product was amplified in a linear range was again determined for each primer set, as described above for housekeeping genes. After hybridization using specific probes (Table II), the mean of the counts of the hybridized probe from six different tonsil donors was calculated for each tonsil B cell subset. Finally, the total number of counts was calculated by summing up the mean counts of each tonsil B cell subset and the fraction of total cDNA in each B cell subset was expressed as a percentage. Rarely, RAG products were detected in the absence of ␤-actin products from single-cell RT-PCR. This occurred in 1–3% of the wells with a detectable RAG product and were omitted from the analysis.

Indirect immunohistochemistry For immunohistochemical staining, 5-␮m cryostat serial sections of unfixed frozen tonsil tissue were prepared. The following Abs were used: biotinylated anti-human IgD (Sigma), followed by alkaline phosphataseconjugated streptavidin (Vector Laboratories, Burlingame, CA), and antihuman RAG2 polyclonal goat Ab (Santa Cruz Biotechnology, Santa Cruz, CA), followed by a peroxidase-conjugated rabbit anti-goat polyclonal Ab (Dako, Carpinteria, CA), followed by a peroxidase-conjugated goat antirabbit polyclonal Ab (Dako). Control sections were stained with a polyclonal preimmunization goat serum with the omission of the anti-IgD Ab. All Abs were diluted in Tris-hydroxymethylaminomethane buffer (pH 7.4) containing 0.5% BSA. Fast Red TR salt (Sigma) was used for detection of Ab-conjugated alkaline phosphatase. Diaminobenzidine (Sigma) was used for detection of Ab-conjugated HRP as described previously (40).

Statistical analysis The differences in RAG1, RAG2, IgG, and IgD mRNA expression in single cells of the various B cell subsets were analyzed using the ␹2 test or Fisher’s test as indicated.

Results RAG1 and RAG2 mRNA expression in tonsil B cell subsets B cell subsets were characterized by surface expression of IgD and CD38 as described previously (32) and identified as follows: naive

The Journal of Immunology

379

Table I. Sequences of oligonucleotides used as primers for the amplification of cDNAa Sequence (5⬘ to 3⬘)

G6PD sense (is) G6PD antisense (is) ␤-actin sense ␤-actin antisense RAG1 sense (is) RAG1 antisense (is) RAG1 nested sense RAG1 nested antisense RAG2 exon 1a sense (is) RAG2 exon 1b sense (is) RAG2 antisense (is) RAG2 nested sense RAG2 nested antisense TdT sense (is) TdT antisense (is) TdT nested sense TdT nested antisense OGG sense OGG antisense VpreB sense (is) VpreB antisense (is) 14.1 sense (is) 14.1 antisense (is) Total IgD sense (is) Total IgD antisense (is) Total IgD nested sense Total IgD nested antisense Total IgG sense (is) Total IgG antisense (is) Total IgG nested sense Total IgG nested antisense a

GTGGAGAATGAGAGGTGGGA GAGAATGAGAGGTGGGATGG GTCCTCTCCCAAGTCCACACA CTGGTCTCAAGTCAGTGTACAGGTAA GAGCAAGGTACCTCAGCCAG AACAATGGCTGAGTTGGGAC TTCTGCCCCAGATGAAATTC TGACCATCAGCCTTGTCCAG GCAGCCCCTCTGGCCTTC GCGGTCTCCAGACAAAAATC TTTCAGACTCCAAGCTGCCT TCTCTGCAGATGGTAACAGTCAG AGCGAAGAGGAGGGAGGTAG ACTTGAGCCCTCGGAAGAAG TTCCCTGCTCCTATGCATTC ATGGCCTCCTCTCCTCAAGA CAAGACCTCACCGAAGTTCG GAGGTGGAGGCTCATCTCAG CTGTAAGGAGGGTGGAGTCG TGCACAGTTGTGGTCCTCAG TCTCCCTCTCCTCCTTCTCC GTAACCCATGGCCTGCTG CGCGTACTTGTTGTTGCTCT GGTACATGGGGACACAGAGC CTGCAGGGGTTAGCAGGTAG ACCGCCAGCAAGAGTAAGAA CTGCAGGGGTTAGCAGGTAG GCTGCCTGGTCAAGGACTAC CATCACGGAGCATGAGAAGA TTCCCCCCAAAACCCAAGGA CATCACGGAGCATGAGAAGA

Intron spanning (is) primers for the initial PCR are marked.

tonsil B cells were IgD⫹ and CD38⫺; pre-GC/GC founder cells were IgD⫹ CD38⫹ and IgD⫹ and CD38⫹⫹; centroblasts were IgD⫺ and CD38⫹⫹; centrocytes were IgD⫺ and CD38⫹; and memory B cells were IgD⫺ and CD38⫺ (Fig. 1). In the six tonsils analyzed, RAG1 mRNA was found in each B cell subset. The greatest quantity of RAG1 mRNA was found in the naive B cell population, which contained about 3 times more than pre-GC/GC founder cells, ⬃8 times more than centroblasts, about 10 times more than centrocytes, and twice as much as memory B cells (Fig. 2). RAG2 mRNA was also prominent in tonsil naive B cells. A comparable amount was found in centrocytes, whereas ⬃20% of the amount was found in centroblasts. No RAG2 mRNA was detected in GC founder cells or memory B cells (Fig. 2). TDT mRNA, OGG, and surrogate light chain mRNA expression in tonsil B cell subsets Expression of TdT mRNA was predominantly found in centroblasts with only minimal expression in naive B cells (Fig. 2). No other subset expressed TdT mRNA. As has been previously noted (41), OGG mRNA was found in equal amounts in CD38⫹⫹ GC founder cells and in centroblasts. Only a marginal amount of OGG mRNA was found in the remaining subsets (Fig. 2). Analysis of VpreB mRNA expression revealed the highest amount in CD38⫹⫹ pre-GC/GC founder cells. They contained about six times more than naive B cells, twice that in CD38⫹ pre-GC/GC founder cells, 1.5 times more than centroblasts, about 2 times more than centrocytes and nearly three times more than memory B cells (Fig. 2). The content of the surrogate (␺) light chain constant region encoding human gene 14.1 mRNA was the highest in centroblasts, which contained 4 times more than naive B cells, 12 times more than CD38⫹ pre-GC/GC founder cells, 3

times more than CD38⫹⫹ pre-GC/GC founder cells, and 22 times more than memory B cells. No expression of 14.1 mRNA was found in centrocytes (Fig. 2). No correlation (r) between the data sets of VpreB and 14.1 was found (r ⫽ 0.452, p ⬎ 0.1). RAG2 expression in tonsil tissue To confirm the expression of RAG protein in tonsil B cells, indirect immunohistochemistry of four different tonsil tissues was performed. GC were visualized by the typical pattern of IgD expression in the mantle zone (Fig. 3A). RAG2 protein-expressing cells were present in the GC (Fig. 3, B and C). A significant number of RAG2-positive cells were also identified in the mantle zone and the interfollicular area. Higher magnification revealed a nuclear distribution of RAG2 with the greatest expression at the periphery of the nuclei (Fig. 3C). Control sections stained with a preimmunization sera were negative (Fig. 3D). Representative consecutive sections of one tonsil are shown.

RAG1, RAG2, TdT, OGG, VpreB, and 14.1 mRNA expression in normal peripheral blood B cells No RAG1 mRNA was found in normal blood B cells isolated from 14 healthy individuals, whereas RAG2 was detected in three samples (Fig. 4). No TdT mRNA expression could be detected in normal peripheral B cells. Eight of 14 peripheral B cell preparations showed low expression of OGG mRNA compared with tonsil mononuclear cells or thymus mononuclear cells. VpreB mRNA expression could be detected in 3 of 14 samples. The 14.1 mRNA was detected at a low level (lane 2) in one sample which did not contain VpreB mRNA (Fig. 4).

380


Table II. Sequences of oligonucleotides used for the detection of PCR products by Southern blotting Sequence (5⬘ to 3⬘)

␤-actin G6PD RAG1 RAG2 TdT VpreB 14.1 OGG IgD IgG

CTCAAGTTGGGGGACAAAAAG AGCTTGAGTCACCTCCTCCA TCTCTGGAGCAATCTCCAGCA TTCCATCTGGATGTAAAGCAT TCCAAAATGAAGACGACCAA CTTGGAACCACAATCCGC AAGGCTACGCTGGTGTGTCT AAACTTTTTCCGGAGCCTGT CCTGATGTGGCTGGAGGACCA GGGTGTACACCTGTGGTTCT

RAG1, RAG2, and IgG mRNA expression by individual B cells of tonsil subsets To analyze RAG1 and RAG2 mRNA expression in greater detail, a very sensitive single-cell RT-PCR method was employed that made it possible to examine the expression of multiple mRNAs by individual phenotypically defined B cells. Analysis of individual tonsil B cells from one patient revealed that 15–26% of individual ␤-actin-positive cells in each subset expressed RAG1 mRNA (Fig. 5A). The differences in RAG1 mRNA expression by the various subsets were not statistically significant ( p ⬎ 0.05, ␹2 test). RAG2 mRNA was frequent in naive, GC founder/pre-GC cells, and memory B cells, but significantly less so in centroblasts and centrocytes ( p ⬍ 0.05, ␹2 test) (Fig. 5A). The frequency of RAG2⫹ cells in centrocytes was significantly greater than in centroblasts ( p ⬍ 0.05, ␹2 test). Exon 1a and exon 1b of the RAG2 gene were used comparably by individual cells of tonsil B cell subsets. Fig. 5B shows the ␤-actin, RAG1, RAG2 exon 1a, and RAG2 exon 1b mRNA expression of CD38⫹IgD⫺ B cells (centrocytes) sorted as individual cells into a 96-well plate. Fifty-four of 96 attempts to sort an individual cell yielded a positive product by single-cell RT-PCR for ␤-actin (56%). Seven of 54 ␤-actin-posi-

FIGURE 1. FACS sorting of tonsil B cell subsets according to their CD38 and IgD surface expression. Tonsil B cell subset tonsils defined by the surface expression of CD19, IgD, and CD38 were sorted with a FACStarPlus. The sort gates for IgD⫹ CD38⫺, IgD⫹ CD38⫹, IgD⫹ CD38⫹⫹, IgD⫺ CD38⫹, and IgD⫺ CD38⫺ cells of tonsil 6 are shown. Sort gates have been set narrowly to reduce contamination from adjacent populations. Different tonsils varied considerably in the absolute numbers of cells in each subset, although the different gated B cell subsets were generally identifiable.

tive wells were also positive for RAG1 and RAG2 mRNA using either RAG2 exon 1a or RAG2 exon 1b (13%). Only the coordinate action of RAG1 and RAG2 together can lead to recombination of the B cell receptor genes (7). Therefore, the frequency of RAG1 and RAG2 mRNA-positive B cells was analyzed as shown in Fig. 5B. RAG1 plus RAG2 expression was found in individual cells in all populations, except centroblasts, which showed no coordinate expression of RAG1 and RAG2 mRNA (Fig. 5A). To assess the stage of maturation of the cells expressing RAG1 plus RAG2 mRNA, IgG mRNA was examined in these individual cells to determine whether they were pre- or postswitch B cells. The frequency of RAG1 plus RAG2 mRNA-positive cells expressing IgG mRNA increased from centrocytes to memory B cells, with 75% of RAG1 plus RAG2 expressing memory B cells also expressing IgG mRNA (Fig. 6). Further analysis of the RAG double-positive cells in the naive and memory B cell subsets in the tonsil was conducted to determine the influence of activation. As can be seen in Fig. 7, the frequency of RAG1 and RAG2 doublepositive cells in activated CD69⫹ memory cells (surface (s) IgD⫺ and CD38⫺) was 10 times higher than in resting CD69⫺ memory B cells. Of note, the RAG1- and RAG2-positive, CD69⫹ memory cells uniformly expressed IgG mRNA (data not shown). As can be

FIGURE 2. RAG1, RAG2, TdT, OGG, and surrogate light chain mRNA expression in tonsil B cell subsets. Tonsil B cell subsets as defined by the surface expression of CD19, IgD, and CD38 were sorted with a FACStarPlus. Specific mRNAs (RAG1, RAG2, TdT, OGG; surrogate light chain) were amplified as described and RT-PCR products were detected by Southern blot hybridization. The amount of specific cDNA detected in each subset was normalized for the content of G6PD cDNA. Each bar represents the mean of counts of hybridized specific probe detected in the subset from six different tonsil donors expressed as a fraction of the total of that cDNA expressed in all of the subsets from the six donors. The brackets indicate SEMs.


381

FIGURE 3. RAG2 expression in human tonsil tissue. Typical GC were identified by anti-IgD immunohistochemical staining of human tonsil tissue (A; original magnification, ⫻100). Serial sections were stained for RAG2 expression (B; original magnification, ⫻100). Frequent RAG2-positive cells were identified in GC and few RAG2-positive cells were found in the mantle zone and in the interfollicular area. Higher magnification revealed the distinct pattern of distribution of RAG2 (C; original magnification, ⫻400). The staining pattern was similar in four different tonsils. Control sections stained with a preimmunization goat serum were negative (D; original magnification, ⫻100).

seen in Fig. 8, no RAG1 and RAG2 double-positive cells were found in the CD69⫺ resting naive B cells (sIgD⫹ and CD38⫺). In contrast, 5% of activated CD69⫹ naive cells expressed RAG1 and RAG2 mRNA (Fig. 8). Of note, these RAG 1- and RAG2-positive, CD69⫹ naive B cells were all positive for IgD mRNA (data not shown). RAG1, RAG2, IgD, and IgG mRNA expression by individual peripheral blood B cells Analysis of peripheral blood B cells of two donors was conducted to investigate whether cells entering the tonsil from the blood expressed RAG1 and RAG2 mRNA. Using single-cell analysis, 3% of CD19- and ␤-actin-positive B cells expressed both RAG1 and RAG2 mRNA (Fig. 9). All of these RAG double-positive cells (n ⫽ 3) expressed IgG mRNA. None of the RAG double-positive cells were positive for IgD mRNA (Fig. 9). Sixteen percent of CD19- and ␤-actin-positive B cells were positive for RAG1 mRNA alone and 26% were positive for RAG2 mRNA alone. Evaluation of the expression of either RAG1 or RAG2 mRNA and the Ig expressed by a particular cell revealed that 5% of ␤-actinpositive cells were positive for only RAG1 and IgG, 9% for RAG2 and IgG, 1% for RAG1 and IgD, and 12% for RAG2 and IgD mRNA. CD19-positive peripheral blood B cells of a second donor were further separated by the expression of sIgD. As shown in Fig. 9, no double RAG1- and RAG2-positive cells were found in the sIgD⫹

subset. However, in the sIgD⫺ subset, 5% of the cells expressed both RAG1 and RAG2 (n ⫽ 4) (Fig. 9). In the IgD⫺ subset, 3 of 4 (75%) of the RAG double-positive cells were positive for IgG mRNA; none was positive for IgD mRNA (Fig. 9). Evaluation of either one of the RAG mRNAs and Ig mRNA expressed by individual cells in the sIgD⫹ subset revealed that 1% of ␤-actin-positive cells were positive for RAG1 and IgG, 1% for RAG2 and IgG, 9% for RAG1 and IgD, and 9% for RAG2 and IgD mRNA. Evaluation of the expression of either one of the RAG mRNAs and the Ig expressed in the sIgD⫺ subset revealed that 12% of ␤-actinpositive cells were positive for RAG1 and IgG, 6% for RAG2 and IgG, 4% for RAG1 and IgD, and 1% for RAG2 and IgD mRNA. In summary, these results have demonstrated that each B cell expressing both RAG1 and RAG2 mRNA in the peripheral blood was positive for IgG mRNA and not IgD mRNA. RAG doublepositive cells in the peripheral blood, therefore, are found in the postswitch memory B cell compartment.

Discussion By using a combination of multiparameter flow cytometric cell sorting and RT-PCR analysis, the expression of RAG mRNA by tonsil and peripheral blood B cell subsets was defined. When bulk preparations of mRNA were analyzed, RAG1 transcripts were detected in all tonsil B cell subsets, whereas RAG2 transcripts were detected in naive B cells, centrocytes, and centroblasts. In contrast, individual B cells of each B cell subset expressed RAG1 and/or

382


FIGURE 4. Expression of RAG1, RAG2 exon 1a, RAG2 exon 1b, TdT, OGG, VpreB, and 14.1 in normal peripheral blood B cells. Peripheral blood B cells of 14 different healthy donors were purified from mononuclear cells by negative selection. RT-PCR products were amplified in the linear range and transferred to a nylon membrane and hybridized with specific ␥-32P-labeled probes. Equal amounts of tonsil and thymus mononuclear cell cDNAs were used as positive controls (lanes 1 and 19, respectively). RT⫺ (lane 17) and no cDNA (lane 18) controls were performed for all samples. Lanes 2–15 are different peripheral B cell mRNA preparations. The Southern blot shown is representative of three consecutive amplifications. Separate PCR amplifications for the alternative exons 1a and 1b of RAG2 were performed.

RAG2 mRNA, with the exception of centroblasts that contained no RAG1 plus RAG2 mRNA-expressing cells. Immunohistochemical staining confirmed that RAG2 protein was expressed not only in the GC, but also in the mantle zone where naive B cells are found. These results indicate a broader distribution of RAG expression in the human tonsil than had been previously reported (30), with RAG found not only in GC B cells, but also in follicular mantle naive cells and post-GC memory B cells. The ability to document RAG mRNA expression in all B cell subsets might be related to the greater sensitivity of the single-cell RT-PCR method compared with the semiquantitative non-nested PCR amplification of bulk cDNA used in the current and previous studies (30). This was apparent since RAG mRNA could be detected in individual cells of subsets that contained no RAG mRNA when the bulk population was analyzed. Moreover, RAG mRNA could be detected in each of these populations when the number of PCR cycles used to amplify bulk cDNA was increased beyond the linear range. These results, therefore, are consistent with the conclusion that RAG mRNA was found in a broader range of tonsil B cell subsets than previously reported (30). One of the unique findings of the current study was the marked discordance in expression of RAG1 and RAG2 mRNA in B cell subsets and, importantly, in individual B cells. The frequency of RAG2 mRNA expressing individual tonsil B cells was 2-fold higher than RAG1 mRNA-expressing cells, indicating that approximately half of the RAG2 mRNA-positive cells were RAG1 mRNA negative. It is unlikely that this discrepancy reflected differential effectiveness of the RAG primers, as careful titrations suggested that each was comparably effective. Moreover, it is unlikely that DNA contamination contributed to the results, as all samples were treated with DNase before analysis and intron span-

FIGURE 5. A, Single-cell mRNA expression of RAG1, RAG2, and IgD in individual tonsil B cell subsets. Individual cells from tonsil B cell subsets defined by the surface expression of CD19, IgD, and CD38 were deposited into 96-well PCR plates with a FACStarPlus. RT-PCR products were detected by Southern blot hybridization. Only ␤-actin mRNA-positive cells (n) were considered for analysis of the frequency of RAG1 and RAG2 mRNA-containing cells. The frequency of single RAG1- or RAG2-positive cells in the various groups as well as the frequency of RAG1 and RAG2 double positive cells are shown on the y-axis from one of the two donors. The second donor revealed similar results. B, Southern blot of individual CD38⫹IgD⫹ tonsil B cells (centrocytes) expressing ␤-actin, RAG1, and RAG2 (exon 1a or alternatively exon 1b). Individual CD38⫹IgD⫺ tonsil B cells (centrocytes) were deposited into 96-well PCR plates with a FACStarPlus. RT-PCR products were detected by Southern blot hybridization. The expression of ␤-actin, RAG1, and RAG2 is shown. RAG2 products were amplified from transcripts alternatively using exon 1a or exon 1b along with exon 2 of RAG2. Depicted are the Southern blot hybridization results from one entire PCR plate. The eight rows of the plate, each consisting of 12 wells, are shown. The upper row of bands indicates cDNA for ␤-actin, the second cDNA for RAG1, the third cDNA for RAG2 using exon 1a primers, and the fourth RAG2 cDNA using exon 1b primers. The arrows indicate individual wells in which both RAG1 and RAG2 cDNA were detected.


383

FIGURE 6. Single-cell mRNA expression of RAG1 and RAG2 in relation to IgD in tonsil B cell subsets. Individual cells from tonsil B cell subsets of two donors were isolated and analyzed as described in the legend of Fig. 5. Only ␤-actin mRNA-positive cells (n) were considered in the analysis. The frequencies of RAG1⫹, RAG2⫹, IgD⫹, and IgG⫹ cells in the RAG1 ⫹ RAG2⫹ subset were determined in the various subsets. The percentages of RAG1- and RAG2-positive cells, IgG⫹ cells, and the IgG⫹ fraction in the RAG1- and RAG2-positive cells are shown.

FIGURE 7. Expression of RAG1 and RAG2 mRNA by individual memory tonsil B cells in relation to surface CD69 expression. Four-color immunostaining of tonsil mononuclear cells was carried out using mAb against CD19, IgD, CD38, and CD69. Individual resting (CD69⫺, n ⫽ 85) and activated (CD69⫹, n ⫽ 70) tonsil memory B cells (CD38⫺, IgD⫺) were deposited in 96-well PCR plates using a FACSVantage flow cytometer, and specific cDNAs were amplified as described. Only ␤-actin mRNA-positive cells (n) were considered in the analysis. The frequencies of RAG1⫹, RAG2⫹, and RAG1 ⫹ RAG2⫹ cells were determined.

ning primers were used. Rather, the data suggest that RAG1 and RAG2 are differentially expressed. Although both RAG1 and RAG2 were down-regulated in individual centroblasts, in other B cell subsets, regulation appeared to be stochastic. RAG mRNA levels are known to be regulated transcriptionally and posttranscriptionally (42– 45). The current data do not permit a conclusion about the mechanisms underlying the discordant expression of RAG1 and RAG2 mRNA in tonsil B cell subsets, but the available data suggest that transcriptional and posttranscriptional effects may play roles. In this regard, coordinate transcription of RAG1 and RAG2 genes has been demonstrated in a murine system using transgenic reporter constructs (46). Even though the regulation was in general coordinate, a considerable number of bone marrow B cells expressed RAG2 promoter activity in the absence of RAG1 expression (46). This differential transcription of RAG2 may contribute to the increased frequency of RAG2 mRNA-expressing cells in the current study. RAG2, and to a lesser extent RAG1, can also be regulated posttranslationally by proliferation-dependent phosphorylation and subsequent degradation of the RAG proteins (43, 44). Currently, there are no available data about whether posttranscriptional regulation of RAG1 and RAG2 mRNA is different. However, in view of the generally coordinate regulation of transcription, it is likely that the discordant expression of RAG1 and RAG2 mRNA levels may reflect differential posttranscriptional regulation. It is notable that the frequency of single RAG-positive cells increased from centrocytes to memory B cells, whereas the frequency of double RAG1- and RAG2-positive individual cells decreased. These results suggest that the capacity for receptor editing may decrease as B cells mature from centrocytes to memory

B cells. Since both the RAG1 and RAG2 enzymes are needed for recombination of the Ig genes, the physiological function of isolated expression of RAG1 or RAG2 mRNA is unknown. To analyze potential influences that might induce coordinate expression of RAG1 and RAG2 mRNA in human B cells, we further analyzed the naive B cell subset in the peripheral blood and the tonsil. Peripheral naive B cells did not show coordinate expression of both RAGs, as was also found for resting CD69⫺ naive B cells in the tonsil. However, a significant number of activated, CD69⫹, naive B cells in the tonsil exhibited coordinate expression of RAG1 and RAG2 mRNA. The absence of RAG double-positive B cells in the CD69⫺ naive tonsil B cell compartment, along with the absence of RAG1 and RAG2 double-positive naive B cells in the peripheral blood, suggest that RAG expression is induced coordinately after naive cells enter the tonsil and become activated, as activation of naive tonsil B cells by Ag is associated with upregulation of CD69 (47). There are only limited and conflicting data on the induction (48, 49) or suppression (30, 49) of RAG expression after cross-linking the B cell receptor to simulate Ag contact by B cells. Induction of RAG in B cells was observed to be induced by sIg engagement (50, 51) and maintained by CD40 engagement (22, 30), possibly by a B cell lineage-specific activator protein-dependent mechanism (52). In contrast, engaging sIg on B cells that are already positive for RAG mRNA down-regulated RAG expression (30). The current results support the hypothesis (23) that activation of naive B cells in secondary lymphoid organs might induce up-regulation of both RAG1 and RAG2 expression following Ag stimulation. Activated CD69⫹ memory B cells contained a 10-fold greater number of coordinate RAG1- and RAG2-expressing cells than

384


FIGURE 8. Expression of RAG1 and RAG2 mRNA by individual naive tonsil B cells in relation to surface cD69 expression. Four-color immunostaining of tonsil mononuclear cells was carried out using mAb against CD19, IgD, CD38, and CD69. Individual resting (CD69⫺, n ⫽ 85) and activated (CD69⫹, n ⫽ 58) naive tonsil B cells (CD38⫺, IgD⫹ were deposited in 96-well PCR plates using a FACSVantage flow cytometer. Only ␤-actin mRNA-positive cells (n) were considered in the analysis. The frequencies of RAG1⫹, RAG2⫹, and RAG1 ⫹ RAG2⫹ cells were determined.

resting memory B cells. This suggests that memory B cells that were generated during a GC reaction decrease their coordinate expression of RAG mRNA once they are removed from activation signals. Alternatively, resting memory B cells that enter the tonsil may up-regulate coordinate RAG1 and RAG2 expression upon Ag-mediated activation. Whichever hypothesis is correct, the regulation of RAG expression in memory B cells appears to be tightly linked to activation status, as it is in naive B cells. The expression of RAG1 and RAG2 mRNA during B cell development in the bone marrow is thought to be tightly controlled and terminated once immature cells express sIgM (2, 42, 44, 45, 53). Two recent reports examining expression of reporter constructs regulated by RAG2 regulatory elements (33, 34) have found that the majority of RAG2 transcription is limited to developing B cells in the bone marrow, but is also found in transitional B cells in the spleen (33). In these studies, no mature murine naive B cells were found to be transcribing RAG genes in vivo, even after immunization (33, 34). In contrast, sIgM⫹sIgD⫹ B cells generated from bone marrow cells in vitro expressed RAG2 after B cell receptor cross-linking with anti-IgM (34). Previously, up-regulation of RAG gene expression was found after Ig receptor crosslinking in both Ig-transgenic and wild-type bone marrow B cells (48). Comparing these studies with the current results suggests that RAG mRNA levels may be extensively regulated in tonsil B cell subsets following engagement of sIg. Part of this regulation may be at the level of transcription. In this regard, marked down-regulation of RAG mRNA in centroblasts followed by up-regulation in centrocytes and memory B cells may be compatible with transcriptional regulation of RAG expression in the tonsil.

FIGURE 9. Single-cell mRNA expression of RAG1, RAG2, IgG, and IgD in peripheral blood B cells. Individual CD19⫹ peripheral blood B cells of one subject and sIgD⫹ and IgD⫺, CD19⫹ B cells of a second donor were deposited into 96-well PCR plates and the presence of RAG1, RAG2, IgG, and IgD mRNA was determined. The y-axis shows the frequency of cells with specific mRNA expressed as a percentage of ␤-actin mRNA⫹ cells (n, CD19⫹ ⫽ 86; n, IgD⫹ ⫽ 87; n, IgD⫺ ⫽ 84). The frequencies of RAG1⫹, RAG2⫹, IgD⫹, IgG⫹, and IgD⫹ and IgG⫹ cells in the RAG1 ⫹ RAG2⫹ cells, calculated as (RAG1 ⫹ RAG ⫹ Ig⫹)/(RAG1 ⫹ RAG2⫹) ⫻ 100, are shown.

It has been suggested that the RAG-expressing cells, previously found in GC, represent immature transitional B cells which have been attracted to the secondary lymphoid organ (19, 22, 23, 25, 33), potentially including the tonsil (29, 30). In contrast, the current results show coordinate RAG1 and RAG2 mRNA expression combined with the expression of IgG in centrocytes and in memory B cells of the tonsil. The IgG primers employed could potentially detect sterile transcripts and, thus, not identify actual postswitch IgG-expressing B cells. However, the complete discordance between the ability to detect IgD mRNA and IgG mRNA suggests that the predominant IgG mRNA species detected represents that obtained from postswitch IgG-expressing B cells, and, therefore, is consistent with the conclusion that RAG1 and RAG2 mRNAs are up-regulated in postswitch IgG-expressing B cells. This conclusion is supported by the analysis of blood B cells, in which RAG1 and RAG2 mRNAs were uniquely found in IgG-expressing cells, even though previous data indicated that resting circulating B cells do not express sterile IgG transcripts (54). In summary, these data, therefore, suggest, that many postswitch B cells generated by GC reactions may express RAG and are not consistent with the contention that RAG is uniquely expressed by transitional cells in secondary lymphoid organs. Surrogate light chain was found to be broadly expressed in all subsets with a maximum in pre-GC/GC founder cells and centroblasts. Therefore, the expression of surrogate light chain was not

The Journal of Immunology correlated with the expression of RAG1 or RAG2 mRNA in the various tonsil B cell subsets, in contrast to previous reports showing coordinate expression (25, 30). These findings would appear to conflict with the conclusion that B cell-specific transcription factors regulate not only the human RAG1 and the murine RAG2 promoter (52, 55, 56), but also surrogate light chain gene expression (52, 55). However, the differences in the RAG1 and RAG2 mRNA expression in comparison to VpreB and 14.1 mRNA expression might again be related to posttranscriptional regulation of the various mRNAs (43, 44). Since primary V(D)J recombination is accompanied by TdTinduced n-nucleotide addition, we investigated whether TdT activity is up-regulated in association with RAG expression. There are disparate data concerning this issue in man and mouse (23, 30). In the current study, expression of TdT mRNA was limited to centroblasts. Previously, TdT mRNA expression was also found in centroblasts, but the predominant expression was reported in centrocytes (30). The different definition of centrocytes in the current (CD38⫹, IgD⫺) and previous (CD38⫹, CD77⫺) studies (30) is unlikely to account for this discrepancy, since CD38⫹, IgD⫺ tonsil B cells were also found to be predominantly CD77low (32). In light of the far broader expression of RAG1 and RAG2 mRNA compared with TdT mRNA levels in tonsil B cell subsets, the current results suggest that if receptor replacement occurs in peripheral lymphoid organs, it is likely to utilize TdT to a limited extent. In summary, the use of single-cell PCR analysis has defined the subsets of tonsil B cells that express RAG mRNA and has clearly shown broad expression after initial activation of naive B cells including centrocytes and postswitch memory B cells that simultaneously express IgG mRNA.

Acknowledgments The technical assistance of Bonnie Darnell, Angie Mobley, Rehana Hussain, Michelle McGuire, Christine Pavlovitch, and Skip Lightfoot is greatly appreciated. We thank Dr. Richard McFarland for providing tissue samples and Dr. Laurie Davis, Dr. Kathryn Meek, Dr. Nancy Farner, Dr. Jisoo Lee, and Dr. Kenji Hayashida for helpful discussion.

References 1. Tonegawa, S. 1983. Somatic generation of antibody diversity. Nature 302:575. 2. Grawunder, U., T. M. J. Leu, D. G. Schatz, A. Werner, A. G. Rolink, F. Melchers, and T. H. Winkler. 1995. Downregulation of RAG1 and RAG2 gene expression in preB cells after functional immunoglobulin heavy chain rearrangement. Immunity 3:601. 3. Difilippantonio, M. J., C. J. McMahan, Q. M. Eastman, E. Spanopoulou, and D. G. Schatz. 1996. RAG1 mediates signal sequence recognition and recruitment of RAG2 in V(D)J recombination. Cell 87:253. 4. Spanopoulou, E., P. Cortes, C. Shih, C.-M. Huang, D. P. Silver, P. Svec, and D. Baltimore. 1995. Localization, interaction, and RNA binding properties of the V(D)J recombination-activation proteins RAG1 and RAG2. Immunity 3:715. 5. Hiom, K., and M. Gellert. 1997. A stable RAG1-RAG2-DNA complex that is active in V(D)J cleavage. Cell 88:65. 6. Schatz, D. G., M. A. Oettinger, and D. Baltimore. 1989. The V(D)J recombination activating gene, RAG-1. Cell 59:1035. 7. Oettinger, M. A., D. G. Schatz, C. Gorka, and D. Baltimore. 1990. RAG-1 and RAG-2, adjacent genes that synergistically activate V(D)J recombination. Science 248:1517. 8. Constantinescu, A., and M. S. Schlissel. 1997. Changes in locus-specific V(D)J recombinase activity induced by immunoglobulin gene products during B cell development. J. Exp. Med. 185:609. 9. Hertz, M., and D. Nemazee. 1998. Receptor editing and commitment in B lymphocytes. Curr. Opin. Immunol. 10:208. 10. MacLennan, I. C. M. 1994. Germinal centers. Annu. Rev. Immunol. 12:117. 11. MacLennan, I. C. M. 1994. From the dark zone to the light. Curr. Biol. 4:70. 12. Kelsoe, G. 1996. The germinal center: a crucible for lymphocyte selection. Semin. Immunol. 8:179. 13. Liu, Y. J., C. Arpin, O. de Bouteiller, C. Guret, J. Banchereau, H. Martinez-Valdez, and S. Lebec. 1996. Sequential triggering of apoptosis, somatic mutation and isotype switch during germinal center development. Semin. Immunol. 8:169. 14. Kroese, F. G. M., A. S. Wubbena, H. G. Seijen, and P. Nieuwenhuis. 1987. Germinal centers develop oligoclonally. J. Immunol. 17:1069.

385 15. MacLennan, I. 1991. The center of hypermutation. Nature 354:352. 16. Ku¨ppers, R., M. Zhao, M.-L. Hansmann, and K. Rajewsky. 1993. Tracing B cell development in human germinal centres by molecular analysis of single cells picked from histological sections. EMBO J. 12:4955. 17. Pascual, V., Y.-J. Liu, A. Magalski, O. de Bouteiller, J. Banchereau, and J. D. Capra. 1994. Analysis of somatic hypermutation in five B cell subsets of human tonsil. J. Exp. Med. 180:32. 18. Berek, C., A. Berger, and M. Apel. 1991. Maturation of the immune response in germinal centers. Cell 67:1121. 19. Liu, Y.-J., and C. Arpin. 1997. Germinal center development. Immunol. Rev. 156:111. 20. Tarlinton, D. 1998. Germinal centers: form and function. Curr. Opin. Immunol. 10:245. 21. Tarlinton, D. 1997. Germinal centers: a second childhood for lymphocytes. Curr. Biol. 7:R155. 22. Hikida, M., M. Mori, T. Takai, K. Tomochika, K. Hamatani, and H. Ohmori. 1996. Reexpression of RAG-1 and RAG-2 genes in activated mature mouse B cells. Science 274:2092. 23. Han, S., B. Zheng, D. G. Schatz, E. Spanopoulou, and G. Kelsoe. 1996. Neoteny in lymphocytes: RAG1 and RAG2 expression germinal center B cells. Science 274:2094. 24. Papavasiliou, F., R. Casellas, H. Suh, X.-F. Qin, E. Besmer, R. Pelanda, D. Nemazee, K. Rajewsky, and M. Nussenzweig. 1997. V(D)J recombination in mature B cells: a mechanism for altering antibody responses. Science 278: 298. 25. Han, S., S. R. Dillon, B. Zheng, M. Shimoda, M. S. Schlissel, and G. Kelsoe. 1997. V(D)J recombinase activity in a subset of germinal center B lymphocytes. Science 278:301. 26. Ohmori, H., and M. Hikida. 1997. Expression and function of recombination activating genes in mature B cells. Crit. Rev. Immunol. 18:221. 27. Liu, Y.-J. 1997. The reuse of B lymphocytes in germinal centers. Science 278: 238. 28. Nemazee, D. 1996. Can receptor editing play an important role in normal B-cell development? J. Autoimmun. 9:259. 29. Giachino, C., E. Padovan, and A. Lanzavecchia. 1998. Re-expression of RAG-1 and RAG-2 genes and evidence for secondary rearrangements in human germinal center B lymphocytes. Eur. J. Immunol. 28:3506. 30. Meffre, E., F. Papavasiliou, P. Cohen, O. de Bouteiller, D. Bell, H. Karasuyama, C. Schiff, J. Banchereau, Y.-J. Liu, and M. Nussenzweig. 1998. Antigen receptor engagement turns off the V(D)J recombination machinery in human tonsil B cells. J. Exp. Med. 188:765. 31. Radic, M. Z., J. Erikson, S. Litwin, and M. Weigert. 1993. B lymphocytes may escape tolerance by revising their antigen receptors. J. Exp. Med. 177:1165. 32. Grammer, A. C., R. D. McFarland, J. Heaney, B. F. Darnell, and P. E. Lipsky. 1999. Expression, regulation and function of B cells expressed CD154 in germinal centers. J. Immunol. 163:4150. 33. Yu, W., H. Nagaoka, M. Jankovic, Z. Misulovin, H. Suh, A. Rolink, F. Melchers, E. Meffre, and M. Nussenzweig. 1999. Continued RAG expression in late stages of B cell development and no apparent re-induction after immunization. Nature 400:682. 34. Monroe, R. J., K. J. Seidl, F. Gaertner, S. Han, F. Chen, J. Sekiguchi, J. Wang, R. Ferrini, L. Davidson, G. Kelsoe, and F. W. Alt. 1999. RAG2:GFP knockin mice reveal novel aspects of RAG2 expression in primary and peripheral lymphoid tissues. Immunity 11:201. 35. Do¨rner, T., S. J. Foster, N. L. Farner, and P. E. Lipsky. 1998. Immunoglobulin ␬ chain receptor editing in systemic lupus erythematosus. J. Clin. Invest. 102:688. 36. Do¨rner, T., C. Heimba¨cher, N. L. Farner, and P. E. Lipsky. 1999. Enhanced mutational activity of V␬ gene rearrangements in systemic lupus erythematosus. Clin. Immunol. 92:188. 37. Brezinschek, H. P., R. I. Brezinschek, and P. E. Lipsky. 1995. Analysis of heavy chain repertoire of human peripheral B cells using single-cell polymerase chain reaction. J. Immunol. 155:190. 38. Testi, R., F. Pulcinelli, L. Frati, P. P. Gazzaniga, and A. Santoni. 1990. CD69 is expressed on platelets and mediates platelet activation and aggregation. J. Exp. Med. 172:701. 39. Mayo, M. J., B. Combes, and R. N. Jenkins. 1996. T-cell receptor V␤ gene utilization in primary biliary cirrhosis. Hepatology 24:1148. 40. Peters, U., T. Papadopoulos, and H. K. Mu¨ller-Hermelink. 1990. MHC class II antigens on lung epithelia of human fetuses and neonates: ontogeny and expression in lungs with histologic evidence of infection. Lab. Invest. 63:38. 41. Kuo, F. C., and J. Sklar. 1997. Augmented expression of a human gene for 8-oxoguanine DNA glycosylase (MutM) in B lymphocytes of the dark zone in lymph node germinal centers. J. Exp. Med. 186:1547. 42. Schatz, D. G., M. A. Oettinger, and M. S. Schlissel. 1992. V(D)J recombination: molecular biology and regulation. Annu. Rev. Immunol. 10:359. 43. Lin, W. C., and S. Desiderio. 1994. Cell cycle regulation of V(D)J recombination-activating protein RAG-2. Proc. Natl. Acad. Sci. USA 91:2733. 44. Lin, W. C., and S. Desiderio. 1995. V(D)J recombination and the cell cycle. Immunol. Today 16:279. 45. Schlissel, M. S., and P. Stanhope-Baker. 1997. Accessibility and the developmental regulation of V(D)J recombination. Semin. Immunol. 9:161. 46. Yu, W., Z. Misulovin, H. Suh, R. R. Hardy, M. Jankovic, N. Yannoutsos, and M. C. Nussenzweig. 1999. Coordinate regulation of RAG1 and RAG2 by cell type-specific DNA elements of 5⬘ of RAG. Science 285:1080. 47. Dono, M., Zupo, S., Augliera, A., Burgio, V. L., Massara, R., Melagrana, A., Costa, M., Grossi, C. E., Chiorazzi, N., and M. Ferrarini. 1996. Subepithelial B

386

48. 49.

50. 51.

52.


cells in the human palatine tonsil. II. Functional characterization. Eur. J. Immunol. 26:2043. Hertz, M., and D. Nemazee. 1997. BCR ligation induces receptor editing in IgM⫹IgD⫺ bone marrow cells in vitro. Immunity 6:429. Sandel, P. C., and J. G. Monroe. 1999. Negative selection of immature B cells by receptor editing or deletion is determined by site of antigen encounter. Immunity 10:289. Hagman, J., and R. Grosschedl. 1994. Regulation of gene expression at early stages of B cell differentiation. Curr. Opin. Biol. 6:222. Ma, A., P. Fisher, R. Dildrop, E. Oltz, G. Rathburn, P. Achacoso, A. Stall, and F. W. Alt. 1992. Surface IgM mediated regulation of RAG gene expression in E mu-N-myc B cell lines. EMBO J. 11:2727. Lauring, J., and M. S. Schlissel. 1999. Distinct factors regulate the murine RAG-2 promoter in B- and T-cell lines. Mol. Cell. Biol. 19:2601.

53. Rolink, A. G., P. Ghia, U. Grawunder, D. Haasner, H. Karasuyama, C. Kalberer, T. H. Winkler, and F. Melchers. 1995. In-vitro analyses of mechanisms of B-cell development. Semin. Immunol. 7:155. 54. Jumper, M. D., J. B. Splawski, P. E. Lipsky, and K. Meek. 1994. Ligation of CD40 induces sterile transcripts of multiple Ig H chain isotypes in human B cells. J. Immunol. 152:438. 55. Verkoczy, L. K., and N. L. Berinstein. 1998. Isolation of genes negatively or positively co-expressed with human recombination activating gene 1 (RAG1) by differential display PCR. Nucleic Acids Res. 26:4497. 56. Verkoczy, L. K., P. A. Marsden, and N. L. Berinstein. 1998. hBRAG, a novel B cell lineage cDNA encoding a type II transmembrane glycoprotein potentially involved in the regulation of recombination activating gene (RAG1). Eur. J. Immunol. 28:2839.