Functional analysis of human memory B-cell subpopulations: IgD+CD27+ B cells are crucial in secondary immune response by producing high affinity IgM

Available online at www.sciencedirect.com R

Clinical Immunology 108 (2003) 128 –137

www.elsevier.com/locate/yclim

Functional analysis of human memory B-cell subpopulations: IgD⫹CD27⫹ B cells are crucial in secondary immune response by producing high affinity IgM Yuhui Shi,a Kazunaga Agematsu,a,b,* Hans D. Ochs,c and Kazuo Suganea a

Department of Infectious Immunology, Shinshu University, Graduate School of Medicine, Asahi 3-1-1, Matsumoto 390-8621, Japan b Department of Pediatrics, Shinshu University, Graduate School of Medicine, Asahi 3-1-1, Matsumoto 390-8621, Japan c Department of Pediatrics, University of Washington, School of Medicine, Seattle, WA 98195, USA Received 19 November 2002; accepted with revision 28 March 2003

The number of memory B cells in peripheral blood has been assayed in various diseases by using CD27 as a memory B-cell marker. However, the defining differences of characteristic and function between the two memory B-cell subpopulations separated by immunoglobulin (Ig)D expression remain to be clearly elucidated. We analyzed here IgD⫹CD27⫹ B cells (circulating B cells 2, cB2) and IgD⫺CD27⫹ memory B cells (cB3) in comparison with IgD⫹CD27⫺ naive B cells (cB1). cB2 were found to be morphologically similar to cB3 with abundant cytoplasm, whereas cB3 expressed CD80, CD86, and CD95 on their surface more predominantly than cB2. A majority of cB2 expressed both IgD and IgM, and cB3 expressed IgA or IgG. Mature ␥1 and ␥2 transcripts were found in cB3, but at very low levels in cB2, and activation-induced cytidine deaminase (AID) mRNA expression was recognized only in cB3. The frequencies of somatic hypermutation in cB2 and cB3 were comparable levels studied by VH5. cB2 did not shift to cB3 in vitro by the stimuli such as via B-cell receptor or CD40. cB2 produced large amounts of IgM predominantly and promptly, which is in accordance with the known characteristics of memory B cells. Taken together, although cB2 are unclass-switched, cB2 have the functions of memory B cells and are not in the process of transition from naive to switched memory B cells, playing a crucial role in secondary immune response by producing high-affinity IgM in the early phase of infections. © 2003 Elsevier Science (USA). All rights reserved. Keywords: B lymphocyte; Memory B cell; IgD; CD27

Introduction The principle of vaccination and immunization for disease prevention depends entirely on the immunological memory that is carried by memory B and T cells [1]. Immunological memory is a defining feature of adaptive immunity and confers the ability to mount more rapid and more robust response to subsequent antigenic encounters [2]. Although central to our understanding of such diverse processes as normal protective immunity, vaccination immunity, transplantation, and autoimmune disease, the mech-

* Corresponding author. Kazunaga Agematsu, Department of Infectious Immunology and Pediatrics, Shinshu University, Graduate School of Medicine, Asahi 3-1-1, Matsumoto 390-8621, Japan. Fax: ⫹81-263-37-3089 E-mail address: [email protected] (K. Agematsu).

anisms underlying the development of both humoral and cellular immune memory responses remain only partially elucidated [3]. Memory B cells, which express CD27 molecule on their surface and undergo somatic hypermutation in immunoglobulin (Ig) variable (V)-region genes, generate Igs rapidly and vigorously in the secondary immune response [4 – 6]. Classically, memory B cells have switched from the initial expression of IgM to that of other Ig classes, resulting in surface expression of IgG, IgA, or IgE, and lack the surface expression of IgD and CD38 [2,7,8]. The recent enormous progress in B-cell analysis clarified that adult circulating B cells could be separated into three subpopulations on the basis of CD27 and IgD expression: IgD⫹CD27⫺ naive B cells (circulating B cell 1, cB1), IgD⫹CD27⫹ B cells (cB2), and IgD⫺CD27⫹ memory B cells (cB3) [4,9 –11]. Klein et al. recently described that

1521-6616/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S1521-6616(03)00092-5

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

IgM⫹IgD⫹CD27⫹ B cells carried somatic hypermutation, indicating that this population is memory cells [12]. Although cB2 presumably represent unclass-switched memory B cells carrying somatically mutated V-region genes and producing IgM and IgG [10 –12], their crucial functions and characteristics are unclear. Recently, analyses of memory B cells in peripheral blood (PB) have been investigated in various disease, such as X-linked hyper-IgM syndrome (XHIM) [13], hyper-IgM syndrome type II (HIGM2) [14], common variable immunodeficiency (CVID) [15,16], human immunodeficiency virus infection [17], primary Sjogren’s syndrome [18], paroxysmal nocturnal hemoglobinuria (PNH) [19], and systemic lupus erythematosus (SLE) [20]. However, not all CD27⫹ B cells are bona fide memory B cells under some conditions of various diseases since CD27 is also an activation antigen of B cells [21]. Since the crucial, functional analysis of memory B-cell subpopulations has not been achieved, it is important to address this issue in normal individuals and patients having immunological disorder. In this study, we investigated the morphology, the expressions of surface molecules, class switch recombination (CSR), frequency of somatic hypermutation, plasma-cell generation, and time course of Ig synthesis among B-cell subpopulations to clarify the situation of these memory B-cell compartments.

129

recombinant IL-2, IL-4, and IL-10 were obtained from Genzyme (Cambridge, MA). Cell preparation PB samples were taken from healthy volunteers after informed consent was given. Mononuclear cells (MNCs) were isolated from PB by Ficoll–Hypaque (Pharmacia, Piscataway, NJ) density gradient centrifugation. The MNCs were separated with 5% sheep erythrocytes into erythrocyte rosette-positive (E⫹) and -negative (E⫺) populations. E⫺ cells were further enriched for B cells by positive selection with anti-CD19 mAb-coated immunomagnetic beads (Dynal, Oslo, Norway). Anti-CD19 mAb was subsequently removed by use of Detach a Bead (Dynal). Highly purified B cells were stained with a combination of IgD–FITC and biotin–CD27, followed by streptavidin–PE before sorting. B-cell proliferation was confirmed as negative in 97% of the population, which reacted with anti-CD20 mAb. No activation was evident in these B cells. CD19⫹ IgD⫹ CD27⫺ (cB1), CD19⫹ IgD⫹ CD27⫹(cB2), and CD19⫹IgD⫺ CD27⫹ B cells (cB3) were isolated from the monocytedepleted E⫺ cells by sorting with a FACStar Plus (Becton Dickinson) under sterile conditions. All three populations thus obtained were ⬎98% purity.

Materials and methods Abs and reagents Anti-CD27 mAb (8H5; IgG1), which does not block the ligation of CD27/CD70, was provided by Dr. T. Morimoto (Dana-Farber Cancer Institute, Boston, MA) [22,23]. FITCconjugated anti-IgD Ab, FITC-conjugated anti-IgG Ab, FITC-conjugated anti-IgA Ab, FITC-conjugated anti-IgM Ab, FITC-conjugated anti-IgE Ab, anti-mouse FITC-conjugated-IgG1, FITC-conjugated anti-CD25 mAb, FITC-conjugated anti-bcl-2 mAb, PE-conjugated anti-CD20 mAb, PE-conjugated anti-CD5 mAb, PE-conjugated anti-CD23 mAb, and PE-conjugated anti-CD38 mAb were purchased from DAKO Japan (Tokyo, Japan). PerCP-conjugated antiCD20 mAb, FITC-conjugated CD25 mAb, FITC-conjugated anti-CD72 mAb, IgD– biotin, and streptavidin–PerCP were purchased from Becton Dickinson (Mountain View, CA) and PE-conjugated anti-CD86 mAb, PE-conjugated anti-CD80 mAb, PE-conjugated anti-HLA-DR mAb, and anti-CD40 mAb (MAB89, IgG1) were purchased from Immunotech (Westbrook, ME). FITC-conjugated anti-Fas (CD95) mAb was purchased from Medical & Biological Laboratories Co., LTD (Watertown, MA). Conjugation of biotin to anti-CD27 mAb was performed by the standard technique using N-hydroxysuccinimido– biotin (Sigma Chemical Co., St. Louis, MO) in our laboratory. Staphylococcus aureus Cowan strain (SAC) and propidium iodide (PI) were obtained from Sigma Chemical Co., and human

Flow cytometric analyses of surface molecules and bcl-2 expression Highly purified B cells were stained with a combination of the mAbs, and two- or three-color analysis of cell surface molecules was performed. For detection of surface IgG, IgA, IgM, and IgD, highly purified B cells before stimulation or after stimulation were blocked by rabbit immunoglobulin fraction (DAKO, Denmark) at room temperature for 20 min. B-cell surface molecules were then stained with a combination of the mAbs. For detection of intracellular bcl-2, cells were fixed with 4% paraformaldehyde in PBS for 15 min at room temperature. After incubation with 0.1% Triton-X ⫹ 0.1% BSA in PBS for 6 min, the cells were washed, followed by the immediate addition of FITC-conjugated antibody and further incubation for 30 min. All these analyses were performed by using a FACScan (Becton Dicknson). Dead cells were removed by staining with PI. Cell fixation CD70 transfectants (CD70T) and CD32 (Fc␥RII) transfectants (CD32T) were prepared as previously described [11,24,25]. The transfectant cells were incubated with 1% paraformaldehyde in PBS for 5 min. After being washed with PBS three times, the cells were cultured in RPMI 1640 ⫹ 10% FCS for 30 min and then used for the analysis.

130

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

Ig assay by ELISA For the IgG, IgM, and IgA syntheses, highly purified cB1, cB2, and cB3 were cultured with SAC plus IL-2 plus IL-10 plus anti-CD40 mAb cross-linked with CD32T (CD40/CD32T). The cells were cultured for 8 days at 37°C in a humidified atmosphere with 5% CO2. The final cell density was 2.5–5 ⫻ 105/ml in a volume of 200 ␮l/well. The plates were coated with goat anti-human Igs (Southern Biotechnology, Birmingham, AL) for the detection of IgG, IgM, and IgA. The cultured supernatants were harvested and added to 96-well flat enzyme-linked immunosorbent assay (ELISA) plates (Nunc). The standard human IgG, IgA, or IgM (Sigma) was also added to the plates. After an overnight incubation, supernatants were discarded and the wells were washed with 0.05% Tween 20 in PBS. Alkaline phosphatase-labeled goat anti-human IgG, IgA, and IgM at a dilution of 1/2500 was added for the detection of IgG, IgA, and IgM, respectively. After 2 h of incubation at room temperature, color detection was performed by 3-[cyclohexylamino]-1-propanesulfonic acid (CAPS) buffer containing p-nitrophenyl phosphate (pNPP) (Sigma). Calibration was performed with PBS at standard zero levels. In this ELISA system, no cross-reaction among IgG, IgA, and IgM occurred. RT-PCR of mature ␥1, ␥2, and AID Highly purified B cells were dissolved in 1 ml TRIzol reagent (Life Technologies, Grand Island, NY). Total RNA was extracted using the acid–guanidinethiocyanate–phenol–chloroform method and then was reverse transcribed into cDNA using oligo(dT) primer (Life Technologies) and Superscript II reverse transcriptase (Life Technologies) in a total volume of 20 ␮l. The following oligonucleotide primers were used for the PCR: for germline C␥1, sense primer 5⬘-ACGAGGAACATGACTGGATGC-3⬘ and antisense primer, 5⬘-TGTGAGTTTTGTCACAAGATTTGGG-3⬘; for germline C␥2, 5⬘-TCTCAGCCAGGACCAAGGAC-3⬘ and 5⬘-ACTCGACACAACATTTGCG-3⬘; for mature C␥1, 5⬘-CCTGGTCACCGTCTCCTCA-3⬘ and 5⬘-TGTGAGTTTTGTCACAAGATTTGGG-3⬘; for mature C␥2, 5⬘-CCTGGTCACCGTCTCCTCA-3⬘ and 5⬘-ACTCGACACAACATTTGCG-3⬘; and for AID, 5⬘-GAGGCAAGAAGACACTCTGG-3⬘ and 5⬘GTGACATTCCTGGAAGTTGC-3⬘. A total of 2 to 5 ␮l cDNA was amplified in PCR using each primer and Taq DNA polymerase (Life Technologies). The amplified products were analyzed on a 1.2% agarose gel containing ethidium bromide and visualized by UV light illumination. The ␤2-microglobulin sense primer 5⬘-GCTATGTGTCTGGGTTTCAT-3⬘ and antisense primer 5⬘ATCTTCAAACCTCCATGATG-3⬘ were used as controls [26]. Sequence analysis of Ig V-region genes VH5 genes were amplified using primers corresponding to the 5⬘ region of the VH5 leader sequence (ATGGGGTCAACCGCCATCCT) and to the 3⬘ C␮ constant region

(GTCCTGTGCGAGGCAGCCAA). Two microliters of cDNA was amplified. The amplification program consisted of 35 cycles of 60 s at 94°C, 60 s at 57°C, and 180 s at 72°C, followed by a final incubation step at 72°C for 5 min. The PCR product was ligated by using the Original TA Cloning Kit (Invitrogen, Carlsbad, CA). Positive colonies were sequenced using the BigDye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA) and an ABI310 sequencer (Applied Biosystems). DNA sequences were compared to the germline VH5 sequence derived from Tomlinson et al. [27]. Statistical analysis Statistical evaluation was performed with Student’s t test. Data were expressed as means ⫾ SD. P ⬍ 0.05 was considered statistically significant.

Results Memory B-cell subpopulations We studied the populations and morphology of circulating human B cells using double-color immunofluorescent staining after high purification. Adult PB B cells were separated into IgD⫹CD27⫺ (cB1), IgD⫹CD27⫹ (cB2), and IgD⫺CD27⫹ (cB3) on the basis of CD27 and IgD expression as shown in Fig 1. As previously reported [11], May– Giemsa staining and flow cytometric analyses revealed the varying morphologies of three of the B-cell populations. cB1 were homogeneously small cells with scant cytoplasm. In marked contrast, cB2 and cB3 were composed of predominantly larger cells with abundant cytoplasm. The expression levels of bcl-2 on three of the B-cell populations in terms of mean fluorescence intensity (MFI) were different. The cytoplasmic staining showed that cB1 displayed 28.3 ⫾ 4.3 MFI of bcl-2 expression, which is lower than that of cB2 (37.6 ⫾ 2.6 MFI, P ⬍ 0.05) or cB3 (44.6 ⫾ 3.6 MFI, P ⬍ 0.01, n ⫽ 3). These data indicate that cB2 and cB3 are similar to each other morphologically and that they may have a longer life span than cB1. Cell surface molecules To clarify the difference between cB2 and cB3, we further examined the expressions of several surface molecules in the B-cell subpopulations (Table 1). CD5 was weakly expressed on cB1 and cB2, but not on cB3. CD23 and CD38 were mainly expressed on cB1. cB1 expressed CD38, although the intensity was at low levels. CD25 expression was not found in B cells. CD72 and HLA-DR were expressed on all of the B-cell populations, yet the expression of CD72 on cB3 was lower than that on cB2. Contrarily, the expression of CD80 on cB3 was higher than that on cB2 and was lacking on cB1 as previously reported [28].

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

131

Fig. 1. Three B-cell subsets separated according to surface IgD and CD27 expression. Peripheral blood E⫺ B cells of healthy donors were stained with FITC-labeled anti-IgD and biotin-labeled anti-CD27 followed by streptavidin–PE. Data are displayed as density plots with green (FITC) fluorescence for IgD and orange (PE) fluorescence for CD27 in logarithmic scale. After sorting, highly purified cB1 (bottom), cB2 (middle), and cB3 (upper) were obtained. May–Giemsa staining (original magnification, ⫻400), cell size displayed with forward scatter (FSC) and side scatter (SSC), and bcl-2 expressions of the three subpopulations (negative control is shown by dot line) are shown. Numbers (shown as means ⫾ SD) indicate the mean fluorescence intensity (MFI) of bcl-2 expression. Data are representative of the results of three experiments using different donors.

Moreover, CD86 and CD95 were expressed predominantly on cB3. These findings reveal that the expressions of some B-cell surface molecules on cB2 and cB3 are distinct. Surface B-cell receptor expression To investigate CSR, we also analyzed the surface IgG (sIgG), sIgA, and sIgM expressions on cB1, cB2, and cB3

Table 1 Expression of surface molecules in cB1, cB2, and cB3a Surface Ag b

CD5 CD23 CD25 CD38 CD72 CD80 CD86 CD95 HLA-DR a

cB1 20.4 30.6 ⬍5 61 96.8 ⬍5 ⬍5 ⬍5 93.8

cB2 ⫾ 4.6 ⫾ 14

c

⫾ 17.5 ⫾ 2.5

⫾ 6.2

11.4 ⬍5 ⬍5 22 79 18.2 ⬍5 6.5 96.7

mRNA expressions of germline/mature ␥1, ␥2, and AID cB3

⫾ 2.3 ⫾ 6.2 ⫾ 10.8 ⫾ 10 ⫾ 2.3 ⫾ 4.7

(Fig. 2A). While about 88 ⫾ 6% of cB1 and 72 ⫾ 4% of cB2 expressed sIgM, only about 15 ⫾ 10% of cB3 expressed sIgM. The IgD⫺IgM⫹CD27⫹ B cells in cB3 are presumably IgM-only memory B cells [29]. In contrast, a greater proportion of cB3 expressed sIgG or sIgA, but not cB2. Expression of sIgE was at very low levels in cB2 and cB3 (data not shown). These data show that cB2 and cB3 express different surface receptors: cB3 express classswitched Ig receptors whereas cB2 have nonswitched Ig receptors.

⬍5 ⬍5 ⬍5 ⬍5 35.3 73.7 26.7 29.1 98.2

⫾ ⫾ ⫾ ⫾ ⫾

15.2 10 3.8 8.9 3.5

Purified B cells were stained with FITC-, PE-, and/or PerCP-labeled mAbs specific for the indicated Ag. b The percentages of positive cells in cB1, cB2, and cB3 are shown. c The percentages are expressed as the means ⫾ SD of five experiments using different donors.

The induction of germline transcripts occurs before CSR. Subsequently, mature transcripts are induced, and Ig synthesis is then initiated. Since cB3, but not cB2, carried class-switched Ig receptors on their surface, we investigated spontaneous germline/mature transcripts in those populations, in addition to that of AID, which is the essential component for CSR [30,31]. As shown in Fig. 2B, both cB2 and cB3 were found to express C␥1 and C␥2 germline transcripts. Although spontaneous expression of mature ␥1 and ␥2 transcripts were found in cB2 and cB3, expression levels of both mature ␥1 and ␥2 in cB2 were lower than those in cB3. No mature ␥1 and ␥2 transcripts were found in cB1. As expected, AID was spontaneously expressed in

132

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

framework regions (FRs) as well as in both complementarity-determining regions (CDRs); 352-nt mutations in 37 clones were detected and the overall frequency was 3.2% (Table 2). Somatic hypermutations in cB3 were also detected in all FRs and CDRs; 533-nt mutations in 40 clones were detected and the overall frequency of mutations in the VH5 genes was 4.53%. No statistical significance was found between the mutation frequencies in cB2 and cB3 (P ⬎ 0.05). We also analyzed ratios of replacement to silent mutations (R/S) in FRs and CDRs of cB2 and cB3. As expected, ratios of R/S in the CDRs of cB2 and cB3 were higher than those in the FRs (Table 2). These findings demonstrate that there is no difference between the levels of somatic hypermutation in cB2 and cB3, indicating that cB2 as well as cB3 produce high-affinity antibodies. Plasma-cell differentiation

Fig. 2. Expression of surface Ig receptors, mature ␥1 and ␥2, and AID transcripts in B-cell subpopulations. (A) Highly purified peripheral blood B cells from five different donors were stained with PE-labeled anti-CD27, biotin-labeled anti-IgD followed by streptavidin–PerCP, and FITC-labeled antibodies specific for the indicated antigen. Expressions of sIgG, sIgM, and sIgA in cB1, cB2, and cB3 were analyzed by flow cytometry. The percentages of surface Ig-positive cells in cB1, cB2, and cB3 are shown. (B) Highly purified cB1, cB2, and cB3 at cell numbers of 0.5 ⫻ 106 cells were used for the extraction of total RNA. RT-PCR was performed as described under Materials and Methods. PCR primers of germline C␥ transcripts were prepared in the initiation region (I␥1 or I␥2) and the hinge of the constant region (C␥1 or C␥2). PCR primers of mature C␥ transcripts were prepared in the region of JH of the V region and of the hinge of the constant region. PCR primers of AID transcripts were prepared including the region of complete cDNA. The ␤2-microglobulin (␤2-MG) was used as a positive control. Data are representative of the results of three experiments using different donors.

cB3, but not in cB2, as well as in cB1. The findings indicate that no class switching process occurs in resting cB2, implying that cB2 are nonswitched B cells. Frequency of somatic mutation in VH regions Somatic hypermutation in Ig V-region genes is the definitive marker of memory B cells. Based on the observation that cB2 underwent somatic hypermutation [12], we next analyzed the difference of somatic mutation frequency between cB2 and cB3. Analysis of the Ig VH5 gene isolated from cB2 revealed that somatic hypermutations were presented in all three

Since predominant differentiation into plasma cells occurs in CD27⫹ B cells [32], we compared the capability of differentiation into plasma cells among cB1, cB2, and cB3 upon various stimuli. Plasma-cell differentiations from cB2 and cB3 were at the same levels without CD27 signaling by CD70 transfectants and were at a lower rate (Fig. 3A). After being stimulated with CD70 transfectants in the presence of SAC plus IL-2 plus IL-10 plus anti-CD40 mAb, both cB2 and cB3 differentiated into plasma cells remarkably at the same rate (Fig. 3B). In contrast, only a small quantity of plasma-cell inductions was recognized in cB1, whether in the presence of CD70 transfectants or not. These data show that cB2 and cB3 have an approximately equipotential ability to differentiate into plasma cells with or without CD27 signaling. Shift from cB2 to cB3 by stimuli The results obtained by the flow cytometric analysis with anti-IgD and anti-CD27 led us to examine whether cB2 can shift to cB3 in vitro. For this purpose, we obtained highly purified adult cB2 by sorting and investigated changes of surface molecules after stimuli. We selected three main signalings: IgD signaling (anti-IgD cross-linked with CD32T), CD40 signaling (CD40/CD32T), and B-cell receptor (BCR) signaling (SAC plus IL-2). The surface expression of BCR such as IgG, IgA, IgM, and IgD did not change significantly upon encountering each of the three types of stimuli (Fig. 4). These data indicate that cB2 cannot differentiate into cB3 at least by stimuli in vitro, suggesting that cB2 are not in the process of differentiating from cB1 to cB3. CD27⫹ B cells generated from adult cB1 or cord blood B cells did not undergo somatic hypermutation after stimulation in vitro [21] and surface expressions of IgG and IgA were not induced (data not shown) despite the remarkable induction of the CD27 expression in vitro. These findings indicate that maybe CD40 engagement or BCR engagement

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

133

Table 2 Somatic hypermutation in cB2 and cB3a Cell population

Donor 1 cB2 cB3 Donor 2 cB2 cB3 Donor 3 cB2 cB3 a b

Number of sequences

Mutations

Total

Numbers

Mutated

Rb

Range

FRs

CDRs

Total

Sb

R/S

Frequencies (%)

FRs

CDRs

FRs

CDRs

Total

Amino acid replacement

13 14

13 (100%) 14 (100%)

1–22 1–17

71 49

39 50

109 99

83 78

26 21

2.9 2.3

2.6 7.3

2.4 1.6

4.3 5.2

2.9 2.4

76 75

12 14

14 (100%) 14 (100%)

2–20 4–23

81 152

44 82

125 234

91 172

36 62

2.1 2.4

3.4 3.5

3 4.8

5.3 8.5

3.5 5.7

78 150

12 12

11 (91.7%) 12 (100%)

0–31 9–35

70 130

48 70

118 200

89 165

29 35

2.3 3.5

5 10.7

2.6 4.8

5.8 8.4

3.3 5.7

73 142

Nucleotide exchange frequency in the VH5 segments of C␮ transcripts. The total number of nucleotides sequenced for the VH5 segments is 294 bp. R and S stand for replacement and silent mutations, respectively. R/S indicate ratios of replacement to silent mutations.

is not sufficient to induce somatic hypermutation and surface expression of class-switched Ig receptors in vitro. Ig production Memory B cells play a crucial role in the secondary immune response by generating Igs rapidly and vigorously. If cB2 are memory B cells producing high-affinity IgM, cB2 may produce IgM promptly by stimuli. Therefore, we examined whether cB2 produce IgM in the early phase as compared with cB1 and cB3. The amount of IgM was determined at different times following stimulation with SAC plus IL-2 plus IL-10 plus CD40/CD32T. Significantly greater production of IgM was measured from cB2 compared with either cB3 or cB1 during the first 3 days following stimulation (Figs. 5A and B). By the 7th day after stimulation, IgM production by cB2 reached a plateau. In contrast, cB3 were the earliest to initiate IgG and IgA syntheses. The findings support a view that cB2 is as important as memory B cells in the secondary immune response by rapidly producing high-affinity IgM. Proportion of cB2 in healthy individuals with age As cB2 play a crucial role in humoral immune response, we further examined the proportion of cB2 in healthy individuals of different ages (Fig. 6). Although the absolute numbers of PB B cells declined with age [33], the percentages of cB2 in PB B cells increased during childhood and adulthood, peaked at about 40 years of age, and then declined in aged persons. The results revealed age-related differences in the percentage and absolute number of cB2 in normal individuals. Discussion In this paper, we investigated the characteristics of a memory B-cell subpopulation in PB, IgD⫹CD27⫹ memory

B cells (referred to in this work as cB2). Our study demonstrated that cB2, which carried somatic hypermutation equally as well as cB3 (switched memory B cells), did not express surface IgG and IgA and did not undergo CSR at the relatively quiescent state. cB2 were prone to differentiate into plasma cells and could produce high-affinity IgM promptly in vitro upon stimuli. The classical criterion of memory B cells is the lack of expression of sIgD and CD38 or the expression of switched Ig isotypes IgG, IgA, and IgE. Functionally, memory B cells proliferate rapidly in response to antigens and produce large amounts of antibodies. The definitive marker of all memory B cells is the presence of somatically mutated high-affinity antigen receptors [2,5,6]. A memory B-cell compartment in PB characterized by cells bearing somatically mutated Vregion genes was recognized in a subset of IgD⫹IgM⫹ B cells expressing CD27 antigen [12]. The findings demonstrated that at least two distinct subpopulations of memory B cells can be identified in PB: IgD⫺CD27⫹ (cB3) and IgD⫹IgM⫹CD27⫹ B cells (cB2). However, the exact characteristics of cB2 have not been sufficiently explored. Morphologically, cB2 and cB3 were larger cells with abundant cytoplasm, in contrast to cB1, probably indicating that both of them were once stimulated upon antigen encounter. The evidence that cB2 and cB3 expressed a greater quantity of bcl-2 than cB1 may indicate a prolonged life span in the memory B cells. However, differences in surface molecule expressions were recognized between cB2 and cB3. Costimulatory molecules CD80 and CD86 and an apoptosis related marker CD95 are activation antigens, with increased expression reported on memory B cells [28,34,36]. In our experiments, their expression levels on cB2 were lower than those on cB3. CD5, CD38, and CD72 are expressed predominantly on naive B cells [34,35,37,38] and their expressions are reduced in the developmental stages from naive B cells to memory B cells [28,34,39]. It is of note that CD38 is weakly expressed on circulating naı¨ve B cells [34], as in our data, but not on some popula-

134

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

Fig. 3. Generation of plasma cells from cB1, cB2, and cB3. (A) cB1, cB2, and cB3 sorted from healthy donors were cultured with medium alone, SAC (0.01%) ⫹ IL-2 (50ng/ml), SAC ⫹ IL-2 ⫹ IL-10 (50ng/ml) ⫹ CD40 (1 ␮g/ml), or SAC ⫹ IL-2 ⫹ IL-10 ⫹ CD40 in the presence of CD70 transfectants (40%) at a final cell density of 10 ⫻ 104 per well for 8 days. The cells were then stained with anti-CD38-FITC and anti-CD20-PE. Ab-coated cells were gated on living cells by cell size and granularity, and dead cells were removed by PI staining and then enumerated by flow cytometric analysis. The percentages of CD38 positive cells are shown. Values are the means ⫾ SD of three independent experiments. (B) After being stimulated with SAC ⫹ IL-2 ⫹ IL-10 ⫹ CD40 in the presence of CD70 transfectants, cB2 and cB3 were stained as described above. Expression of CD38 and CD20 on B cells are shown on a log scale. The percentages of CD38-positive cells are shown.

tions of tonsillar naı¨ve B cells [6]. The expression levels of CD5, CD38, and CD72 on cB2 were higher than those on cB3. Analysis by the expression of surface molecules reveals that cB2 and cB3 are two of independent, different subpopulations of memory B cells, but cell surface marker analysis shows that cB2 are rather more similar to cB1 than they are to cB3. CSR of the Ig genes takes place in germinal center (GC) [40]. CSR replaces the Ig heavy chain constant region (CH) gene to be expressed from C␮ to other CH genes, resulting in switch of the Ig isotype from IgM to IgG, IgA, or IgE. B-cell Ig synthesis is preceded by transcription of the germline mRNA, and mature mRNA is expressed after CSR. In the view of classical immunological memory, Ig isotype switching has occurred in memory B cells [41]. Since cB2 express both IgD and IgM, and cB3 do not express IgD, it

is suggested that cB2 are unclass-switched and cB3 are class-switched memory B cells [12]. To confirm this issue, we investigated expression levels of sIgG, sIgA, and spontaneous mature ␥1, ␥2 transcripts by using RT-PCR in both memory subpopulations. We found here that sIgG/sIgA were expressed in cB3, but not in cB2, and mature ␥1, ␥2 transcripts were expressed in cB3, but only very faintly in cB2. AID is the only B-cell-specific factor required for initiation of the CSR [30,31]. AID also was spontaneously expressed in cB3, but was not found in cB2. These findings support the view that cB3 are class-switched cells and cB2 are nonswitched cells. Analysis of the mutational status of Ig V-region genes revealed that cB1 expressed unmutated Ig genes [21]. Both cB2 and cB3 carried somatic hypermutation equally, and mutations were presented in all CDRs and FRs of Ig VH5 genes in this study (data not shown). cB2 and cB3 showed a level of somatic hypermutation in their V-region genes that was in the same range as that previously reported in memory B cells [5,12,29,42,43]. Memory B cells have been selected within GCs for antibody expression and high-affinity antigen binding, it is expected that R mutations should be counterselected within the FRs (i.e., low R/S), whereas the antigen-binding CDRs should be selected for affinityincreasing R mutations (i.e., high R/S). The higher R/S within CDRs of cB2 and the presence of somatic mutations in Ig V-region genes in nonswitched cB2 as reported by Klein et al. [12] confirm that these B cells are indeed memory B cells and can produce high-affinity antibody. Somatic hypermutation of Ig V-region genes and CSR occurs within GCs in secondary lymphoid organs during the differentiation from naive into memory B cells. However, somatic hypermutation and isotype switch are two independent processes and somatic mutation can occur within GCs

Fig. 4. Expressions of surface Igs on cB2 after stimulation. Highly purified cB2 (10 ⫻ 104/well) were cultured with IL-2 (50 ng/ml) ⫹ CD40 (1 ␮g/ml)/CD32T (40%), anti-IgD (1 ␮g/ml) with CD32T (40%), or SAC (0.01%) ⫹ IL-2 (50 ng/ml) for 8 days. The cells were then stained with FITC-labeled mAbs specific for the indicated surface Ig. The antibodycoated cells were gated on living cells by cell size and granularity and finally counted by means of flow cytometric analysis. Dead cells were removed by PI staining. Data are representative of the results of four experiments using different donors. Results are expressed as means ⫾ SD.

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

Fig. 5. Production of IgM, IgG, and IgA by cB1, cB2, and cB3 after stimulation. After being sorted, highly purified cB1, cB2, and cB3 (10 ⫻ 104/well) were cultured with SAC (0.01%) ⫹ IL-2 (50 ng/ml) ⫹ IL-10 (50 ng/ml) ⫹ CD40 (1 ␮g/ml)/CD32T (40%) for 8 days. (A) Supernatants were collected daily following stimulation. IgM, IgG, and IgA production were measured by ELISA as described under Materials and Methods. Amounts of IgM, IgG, and IgA production by cB1, cB2, and cB3 at various times after stimulation are shown. Similar results were obtained in two independent experiments. (B) After being cultured daily, supernatants were collected, and then fresh medium was added for the next 1-day culture. Amounts of IgM, IgG, and IgA production by cB1, cB2, and cB3 per day with the passage of time were detected. Data are representative of the results of two experiments.

without triggering isotype switch [21,40]. Isotype switch is triggered at the centrocyte after somatic hypermutation is initiated within centroblasts. It has been demonstrated that B cells, carrying somatic mutations without CSR, reside in tonsils [6]. It is consistent with our observations in PB that cB2 carry somatic mutations and do not undergo CSR. In our in vitro stimulation system, no increased expressions of sIgG and sIgA were determined in cB2 after stimuli, and sIgD did not disappear. This result demonstrates that cB2 cannot differentiate into cB3, at least in vitro, by the conventional stimuli. To produce antibodies, the differentiation of B cells into the specific antibody-secreting cells, plasma cells, is required. Triggering via B-cell Ig receptors by antigens, cytokines such as IL-2 and IL-10, and direct cell-to-cell contact between T and B cells play an important role in the differentiation into plasma cells. Memory B cells differentiated into plasma cells rapidly by activated helper T cells via CD27/CD70 in the presence of several cytokines such as IL-10 and IL-2 [32]. A considerable number of memory B cells, both cB2 and cB3, but not naive B cells, cB1, could

135

differentiate into plasma cells upon stimulation equally, with CD27 signaling in the memory B cells enhancing the generation of plasma cells. The findings also support the view that cB2 is a memory B cell like cB3. CD27 is crucial in controlling the differentiation of memory B cells into plasma cells and the expression of CD27 on memory B cells is important for the prompt differentiation into plasma cells. Generating Igs rapidly and vigorously in the secondary immune response is another important feature of memory B cells. Upon stimulation in vitro, cB3 produced large amounts of Igs whereas cB2 produced IgM predominantly (Fig. 5 and Refs. 10,11,44). Alternatively, the functional difference between the two memory B-cell subpopulations may be reflected in the time course. Since cB2 produced large amounts of IgM earlier upon stimulation than cB1 and cB3, cB2 may produce high-affinity antigen-specific IgM early during infection. Weller et al. [45] reported that cB2 in XHIM carried mutated Ig sequence and that cB2 could generate fast protective responses in humans to T-independent antigens carried by infectious agents such as bacterial polysaccharides and viral repetitive surface determinants [46,47]. The decline of cB2 numbers in children and elderly persons demonstrates a poor secondary humoral immunity by IgM and may be reflected in the susceptibility to infections in these populations. A remarkable decrease of cB2 and cB3 was reported in SLE [20], PNH [19], and human immunodeficiency virus infection [17]. Memory B-cell compartment deficiency in cB3, but not in cB2, was found in various immunodeficiency diseases such as CVID [15,16] and XHIM [13]. The patients with HIGM2 had large amounts of cB2 but did not carry apparent somatic hypermutation (Ref. 14 and our unpublished data). Also, cB2 increased in some of CVID patients did not carry somatic hypermutation in our experiments [15]. The cB2 found in

Fig. 6. Distribution of cB2 in healthy persons. Correlation between age and percentage of cB2 in B cells is shown. MNCs from cord blood and PB of healthy persons (n ⫽ 140) aged 0 –99 years were stained with anti-IgD– FITC, anti-CD20 –PerCP and anti-CD27– biotin followed by streptavidin– PE. Three-color analysis was carried out by gating CD20-positive B cells. The percentages of cB2 are shown as means ⫾ SD.

136

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

such abnormal conditions may be, at least in part, activated B cells. Thus, IgD⫹CD27⫹ B cells in PB are an independent, functional subpopulation of memory B cells and may play a crucial role in secondary immune response by their prompt synthesis of high-affinity IgM. The findings demonstrated here provide a new aspect in protective humoral immunity. In addition, this new approach to the classification of circulating B cells by the expression of IgD and CD27 may provide further clues for analysis of the morbid state in various diseases caused by immunological disorder.

[16]

[17]

[18]

References [19] [1] J. Spren, T and B memory cells, Cell 76 (1994) 315–322. [2] D. Gray, Immunological memory, Annu. Rev. Immunol. 11 (1993) 49 –77. [3] R. Dutton, S. Swain, L. Bradley, The generation and maintenance of memory T and B cells, Immunol. Today 20 (1999) 291–293. [4] K. Agematsu, S. Hokibara, H. Nagumo, A. Komiyama, CD27: a memory B-cell marker, Immunol. Today 21 (2000) 204 –206. [5] U. Klein, T. Goossens, M. Fischer, H. Kanzler, A. Braeuninger, K. Rajewsky, R. Kuppers, Somatic hypermutation in normal and transformed human B cells, Immunol. Rev. 162 (1998) 261–280. [6] V. Pascual, Y. Liu, A. Maglski, O. de Bouteiller, J. Banchereau, J. Capra, Analysis of somatic mutation in five B cell subsets of human tonsil, J. Exp. Med. 180 (1994) 329 –339. [7] Y.J. Liu, C. Barthelemy, O. de Bouteiller, C. Arpin, I. Durand, J. Banchereau, Memory B cells from human tonsils colonize mucosal epithelium and directly present antigen to T cells by rapid up-regulation of B7-1 and B7-2, Immunity 2 (1995) 239 –248. [8] J. Banchereau, F. Rousset, Human B lymphocytes: phenotype, proliferation, and differentiation, Adv. Immunol. 52 (1992) 125–262. [9] D. Maurer, W. Holter, O. Majdic, G.F. Fisher, W. Knapp, CD27 expression by a distinct subpopulation of human B lymphocytes, Eur. J. Immunol. 20 (1990) 2679 –2684. [10] D. Maurer, G. Fischer, I. Fae, O. Majdic, K. Stuhlmeier, N. Von Jeney, W. Holter, W. Knapp, IgM and IgG but not cytokine secretion is restricted to the CD27⫹ B lymphocyte subset, J. Immunol. 148 (1992) 3700 –3705. [11] K. Agematsu, H. Nagumo, F.C. Yang, T. Nakazawa, K. Fukushima, S. Ito, K. Sugita, T. Mori, T. Kobata, C. Morimoto, A. Komiyama, B cell subpopulations separated by CD27 and crucial collaboration of CD27⫹ B cells and helper T cells in immunoglobulin production, Eur. J. Immunol. 27 (1997) 2073–2079. [12] U. Klein, K. Rajewsky, R. Kuppers, Human immunoglobulin (Ig)M⫹IgD⫹ peripheral blood B cells expressing the CD27 cell surface antigen carry somatically mutated variable region genes: CD27 as a general marker for somatically mutated (Memory) B cells, J. Exp. Med. 188 (1998) 1679 –1689. [13] K. Agematsu, H. Nagumo, K. Shinozaki, S. Hokibara, K. Yasui, K. Terada, N. Kawamura, T. Toba, S. Nonoyama, H.D. Ochs, A. Komiyama, Absence of IgD⫺CD27⫹ memory B cell population in X-linked hyper-IgM syndrome, J. Clin. Invest. 102 (1998) 853– 860. [14] P. Revy, T. Muto, Y. Levy, F. Geissmann, A. Plebani, O. Sanal, N. Catalan, M. Forveille, R. Dufourcq-Labelouse, A. Gennery, I. Tezcan, F. Ersoy, H. Kayserili, A. Ugazio, N. Brousse, M. Muramatsu, L. Notarangelo, K. Kinoshita, T. Honjo, A. Fischer, A. Durandy, Activation-induced cytidine deaminase (AID) deficiency causes the autosomal recessive form of the Hyper-IgM syndrome (HIGM2), Cell 102 (2000) 565–575. [15] K. Agematsu, T. Futatani, S. Hokibara, N. Kobayashi, M. Takamoto, S. Tsukada, H. Suzuki, S. Koyasu, T. Miyawaki, K. Sugane, A.

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

Komiyama, H. Ochs, Absence of memory B cells in patients with common variable immunodeficiency, Clin. Immunol. 103 (2002) 34 – 42. K. Warnatz, A. Denz, R. Draeger, M. Braun, C. Groth, G. WolffVorbeck, H. Eibel, M. Schlesier, H. Peter, Severe deficiency of switched memory B cells (CD27(⫹)IgM(⫺)IgD(⫺)) in subgroups of patients with common variable immunodeficiency: a new approach to classify a heterogeneous disease, Blood 99 (2002) 1544 –1551. H. Nagase, K. Agematsu, K. Kitano, M. Takamoto, Y. Okubo, A. Komiyama, K. Sugane, Mechanism of hypergammaglobulinemia by HIV infection: circulating memory B-cell reduction with plasmacytosis, Clin. Immunol. 100 (2001) 250 –259. A. Hansen, M. Odendahl, K. Reiter, A. Jacobi, E. Feist, J. Scholze, G. Burmester, P. Lipsky, T. Dorner, Diminished peripheral blood memory B cells and accumulation of memory B cells in the salivary glands of patients with Sjogren’s syndrom, Arthritis Rheum. 46 (2002) 2160 –2171. S. Richards, G. Morgan, P. Hillmen, Immunophenotypic analysis of B cells in PNH: insights into the generation of circulating naive and memory B cells, Blood 96 (2000) 3522–3528. M. Odendahl, A. Jacobi, A. Hansen, E. Feist, F. Hiepe, G. Burmester, P. Lipsky, A. Radbruch, T. Dorner, Disturbed peripheral B lymphocyte homeostasis in systemic lupus erythematosus, J. Immunol. 165 (2000) 5970 –5979. H. Nagumo, K. Agematsu, N. Kobayashi, K. Shinozaki, S. Hokibara, H. Nagase, M. Takamoto, K. Yasui, K. Sugane, A. Komiyama, The different process of class switching and somatic hypermutation: a novel analysis by CD27⫺ naive B cells, Blood 99 (2002) 567–575. T. Kobata, S. Jacquot, S. Kozlowski, K. Agematsu, SF. Schlossman, C. Morimoto, CD27-CD70 interactions regulate B-cell activation by T cells, Proc. Natl. Acad. Sci. USA 92 (1995) 11249 –11253. K. Sugita, T. Hirose, D.M. Rothstein, C. Donahue, S.F. Schlossman, C. Morimoto, CD27, a member of the nerve growth factor receptor family, is preferentially expressed on CD45RA⫹ CD4 T cell clones and involved in distinct immunoregulatory functions, J. Immunol. 149 (1992) 3208 –3216. G. Peltz, M. Trounstine, K. Moore, Cloned and expressed human Fc receptor for IgG mediates anti-CD3 dependent lymphoproliferation, J. Immunol. 141 (1988) 1891–1896. H. Karasuyama, A. Kudo, F. Melchers, The proteins encoded by the VpreB and ␭ 5 pre-B cell-specific genes can associate with each other and with ␮ heavy chain, J. Exp. Med. 172 (1990) 969 –972. D. Gussow, R. Rein, I. Ginjaar, F. Hochstenbach, G. Seema, A. Kottman, H. Ploegh, The human b2-microglobulin gene: primary structure and definition of the transcription unit, J. Immunol. 139 (1987) 3132–3138. I. Tomlinson, G. Walter, J. Marks, M. Llewelyn, G. Winter, The repertoire of human germline VH sequences reveals about fifty groups of VH segments with different hypervariable loops, J. Mol. Biol. 227 (1992) 776 –798. A. Bar-Or, E. Oliveira, D. Anderson, J. Krieger, M. Duddy, K. O’Connor, D. Hafler, Immunologicaal memory: contribution of memory B cells expressing costimulatory molecules in the resting state, J. Immunol. 167 (2001) 5669 –5677. U. Klein, R. Kuppers, K. Rajewsky, Evidence for a large compartment of IgM-exprssing memory B cells in humans, Blood 89 (1997) 1288 –1298. M. Muramatsu, K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, T. Honjo, Classes switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potentiao RNA editing enzyme, Cell 102 (2000) 553–563. I. Okazaki, K. Kinoshita, M. Muramatsu, K. Yoshikawa, T. Honjo, The AID enzyme induces class switch recombination in fibroblasts, Nature 727 (2002) 1–5. K. Agematsu, H. Nagumo, Y. Oguti, T. Nakazawa, K. Fukushima, K. Yasui, S. Ito, T. Kobata, C. Morimoto, A. Komiyama, Generation of

Y. Shi et al. / Clinical Immunology 108 (2003) 128 –137

[33]

[34]

[35]

[36]

[37] [38] [39] [40]

plasma cells from peripheral blood memory B cells: synergistic effect of IL-10 and CD27/CD70 interaction, Blood 91 (1998) 173–180. C. Franceschi, D. Monti, P. Sansoni, A. Cossarizza, The immunology of exceptional individuals: the lesson of centenarians, Immunol. Today 16 (1995) 12–16. J. Bohnhorst, M. Bjorgan, J. Thoen, J. Natvig, K. Thompson, Bm1Bm5 classification of peripheral blood B cells reveals circulating germinal center founder cells in healthy individuals and disturbance in the B cell subpopulation in patients with primary Sjogren’s syndrome, J. Immunol. 167 (2001) 3610 –3618. A. Grammer, R.D. McFarland, J. Heaney, B.F. Darnell, P.E. Lipsky, Expression, regulation, and function of B cell-expressed CD154 in germinal centers, J. Immunol. 163 (1999) 4150 – 4159. S. Tangye, Y. Liu, G. Aversa, J. Phillips, J. de Vries, Identification of functional human splenic memory B cells by expression of CD148 and CD27, J. Exp. Med. 188 (1998) 1691–1703. P. Youinou, C. Jamin, P. Lydyard, CD5 expression in human B-cell populations, Immunol. Today 20 (1999) 312–326. T. Tsubata, Co-receptors on B lymphocytes, Curr. Opin. Immunol. 11 (1999) 249 –255. T. Kipps, The CD5 B cell, Adv. Immunol. 47 (1989) 117–185. Y.J. Liu, F. Malisan, O. de Bouteiller, C. Guret, S. Lebecque, J. Banchereau, F.C. Mills, E.E. Max, H. Martinez-Valdez, Within ger-

[41] [42]

[43]

[44]

[45]

[46] [47]

137

minal centers, isotype switching of immunoglobulin genes occurs after the onset of somatic mutation, Immunity 4 (1996) 241–250. J. Banchereau, F. Rousset, Human B lympyocytes: phenotypes, proliferation, and differentiation, Adv. Immunol. 52 (1992) 125–162. H. Brezinsche, S. Foster, T. Dorner, R. Brezinschek, P. Lipsky, Pairing of variable heavy and variable chains in individual naive and memoty B cells, J. Immunol. 160 (1998) 4762– 4767. M. Fischer, U. Klein, R. Kuppers, Molecular single-cell snalysis reveals that CD5-positive peripheral blood B cells in healthy humans are characterized by rearranged Vkappa genes lacking somatic mutation, J. Clin. Invest. 100 (1997) 1667–1676. T. Defrance, B. Vanbervliet, F. Briere, I. Durand, F. Rousset, J. Banchereau, Interleukin 10 and transforming growth factor beta cooperate to induce anti-CD40-activated naive human B cells to secret immunoglobulin A, J. Exp. Med. 175 (1992) 671– 682. S. Weller, A. Faili, C. Garcia, M. Braun, F. Deist, G. de Saint Basile, O. Hermine, A. Fischer, C. Reynaud, J. Weill, CD40-CD40L independent Ig gene hypermutation suggests a second B cell diversification patyway in humans, Proc. Natl. Acad. Sci. USA 98 (2001) 1166 –1170. M. Bachmann, P.S. Ohashi, The role of T-cell receptor dimerization in T-cell activation, Immunol. Today 20 (1999) 568 –576. J. Mond, A. Lees, C. Snapper, T cell-independent antigens type 2, Annu. Rev. Immunol. 13 (1995) 655– 692.