Detailed characterization of Ag-specific naive and memory B cell Ab repertoires elucidates the molecular basis for the generation of Ab diversity and the optimization of Ab structures that bind microbial Ags. In this study, we analyzed the immunophenotype and VH gene repertoire of rotavirus (RV) VP6-specific B cells in three circulating naive or memory B cell subsets (CD19+IgD+CD27, CD19+IgD+CD27+, or CD19+IgDCD27+) at the single-cell level. We aimed to investigate the influence of antigenic exposure on the molecular features of the two RV-specific memory B cell subsets. We found an increased frequency of CD19+IgD+CD27+ unclass-switched memory B cells and a low frequency of somatic mutations in CD19+IgDCD27+ class-switched memory B cells in RV-specific memory B cells, suggesting a reduced frequency of isotype switching and somatic mutation in RV VP6-specific memory B cells compared with other memory B cells. Furthermore, we found that dominance of the VH1–46 gene segment was a prominent feature in the VH gene repertoire of RV VP6-specific naive B cells, but this dominance was reduced in memory B cells. Increased diversity in the VH gene repertoire of the two memory B cell groups derived from broader usage of VH gene segments, increased junctional diversity that was introduced by differential TdT activities, and somatic hypermutation.

Rotavirus (RV)3 is the most common cause of dehydrating diarrhea in infants and young children worldwide, accounting for approximately one-third of cases of severe diarrhea requiring hospitalization (1). RV infects and replicates in the epithelial cells of the small intestine (2). The mechanisms responsible for immunity to RV infection and illness are not completely understood, especially in humans. B cells likely contribute to immunity through Th cell-dependent secretion of RV-specific IgA and IgG, while CTL play a role in the clearance of the virus. B cell-derived Ig appears to play a more important role in the protective RV-specific memory response (3, 4, 5, 6). The detection of robust T cell-independent B cell responses in early RV infection in mice implicated B cells as a major determinant of initial RV clearance (7). Furthermore, naive B cells also were shown to be the principal APCs associated with induction of intestinal IgA production after s.c. RV injection in mice (8).

Primary infection by RV induces production of RV-specific memory B and T cells, which reduce severity of disease, but this response is not usually sufficient to prevent reinfection by the virus completely (4, 5, 6). Partial immunity also is observed following immunization with RV vaccines, including the most recently licensed vaccine, Rotateq, which induces ∼70% protection against infection after administration of three doses (9, 10). The molecular basis for the limited immunity against RV is still unknown. A better understanding of the protective mechanisms against RV infection will facilitate design of improved vaccine and therapeutic strategies against RV infection and disease.

RV is a dsRNA virus with 11 segments of RNA genome packed in a triple-layered viral capsid. Three of the RV structural proteins, VP4, VP6, and VP7, have important antigenic properties. The intermediate-layer capsid protein, VP6, is the most antigenic protein of RV and mediates group and subgroup specificity, while the outer-layer proteins, VP4 and VP7, mediate serotype P and serotype G specificities, respectively (11). VP6 does not induce classically neutralizing Abs; however, VP6-specific polymeric IgA molecules are protective in vivo, likely through inhibition of the transcriptional machinery of RV inside the cell during transcytosis of IgA (12, 13). Previous studies have shown that the RV VP6 protein interacted with a large fraction of naive B cells from both human adult and neonatal blood samples, suggesting that the naive B cell-VP6 interaction might influence the strength and quality of the acquired immune response (14).

Our previous studies have shown that VH gene segment usage in VP6-specific human B cells was distinct from that of randomly selected B cells, and was similar in both adults and infants (15, 16). VH1–46 is the immunodominant Ig gene segment used in both circulating and intestinal-homing VP6-specific B cells (17). However, whether the VH gene bias resides in the naive and/or in the memory VP6-specific B cell compartment remains to be determined. Recent progress in lymphocyte-phenotyping methods has facilitated better definition of memory B cell subsets in humans. Circulating human B cells can be separated into at least three major naive or memory B cell groups: CD19+IgD+CD27 naive B cells, CD19+IgD+CD27+ unclass-switched memory B cells and CD19+IgDCD27+ class-switched memory B cells (18, 19). Among these groups of B cells, CD19+IgD+CD27+ unclass-switched memory B cells are the most recently recognized population. Their unique characteristics suggest a particular role in T cell-independent B cell responses, Ab responses against encapsulated bacteria, and marginal zone B cell responses (20, 21). The molecular and cellular basis for development of diverse and effective Ab repertoires in these two memory B cell subsets remains only partially elucidated.

In this study, we investigated the Ab repertoire in three human circulating naive and memory B cell subsets, in cells that are specific for RV. Using GFP-labeled RV VP2/6 double-layered particles, we identified and simultaneously isolated the RV VP6-specific circulating naive and memory B cells from healthy adult volunteers. Individual variable genes of Ig H chains from single cells were cloned and analyzed. Consistent with previous reports from our group, we found that dominance of the VH1–46 gene segment was a prominent feature in the Ab repertoire of all RV-specific B cell groups. VH1–46 dominance, however, was reduced in the Ab repertoire of RV-specific memory B cells compared with that of RV-specific naive B cells. Diversity in the Ab repertoire of the two memory B cell groups was increased compared with that of naive cells, as evidenced by both broader usage of VH gene segments and increased number of somatic mutations.

Peripheral blood samples (n = 10) from healthy adult volunteers, aged 20–45 years, were used for study. All samples were obtained following informed consent under approval from the Vanderbilt University Medical Center Institutional Review Board.

We isolated single human RV VP6-specific B cell clones from peripheral blood samples using single-cell flow cytometric sorting. RV-specific B cells were identified by staining with GFP-labeled virus-like particles (VLP). For this study, we used GFP-labeled RV VP2/6 double-layered particles (DLP). GFP-VLPs for single-cell sorting were produced by coinfection of Sf9 insect cells with recombinant baculoviruses, as described (22). These particles exhibit the RV VP6 protein on the outside surface in a conformationally correct fashion and contain the GFP molecule that is used for fluorescence detection fused to the RV VP2 protein on the inside (unexposed portion) of the particle. These particles were carefully titrated with RV VP6-specific murine B cell hybridoma cells and were shown to bind RV VP6-specific murine B cell hybridoma cells in a specific manner. The origin of VP6 was the RV wild-type strain RF. PBMCs were isolated from blood samples by Ficoll-Hypaque density gradient centrifugation, then stained for 30 min at 4°C in the dark using GFP-labeled RV VP2/6 DLP and fluorescent-conjugated mouse anti-human Abs, including anti-CD19-PE-Cy7, anti-IgD-PE, anti-CD27-allophycocyanin, anti-CD3/CD14-allophycocyanin-Cy7 (BD Biosciences). Cells were processed immediately for flow cytometric analysis and cell sorting using a FACSAria cytometer (BD Biosciences). Cells expressing CD3 or CD14 (T cell or monocyte markers) were excluded from sorting. After each experiment, a portion of the sorted sample was analyzed to determine the postsort purity.

We used a culture system as previously described for the expansion of single B cells into clones (22). Briefly, single B cells were incubated with recombinant human IL-2 and IL-4, supernatant from mitogen-stimulated primary human T cells, and irradiated fibroblasts persistently transfected with a plasmid expressing the ligand of CD40 (human CD154). Before RT-PCR for recovery of Ab genes, specificity of the B cell clone for RV was confirmed by human Ig capture ELISA of Abs in culture supernatants using RV DLP as Ag. The DLP used in this ELISA differed from those used in sorting, in that they were naturally occurring DLP isolated from MA104 cells infected with the rhesus rotavirus strain of RV (designated RRV). The rationale for using this Ag in the ELISA screening was to avoid detection of clones secreting Abs that were not RV VP6 specific (such as GFP specific), if such clones were selected inadvertently in the flow cytometric sorting. Previously, our group confirmed that these criteria for selection of RV-specific cells were effective, by generating rAbs containing the immunodominant VH gene segments and showing that they bound to RV-infected cells or purified RV particles. We found these purified Fabs bound in a dose dependent and specific manner to DLP, but not to complete (triple-layered) RRV particles, showing again their specificity for VP6.

We used a human Ig capture ELISA for determining the isotype of the Abs produced by single B cell-derived clones. Unconjugated goat anti-human Ig (H&L) Ab (Southern Biotechnology Associates), diluted 1/1000 in sodium carbonate buffer was used as capture Ab. Alkaline phosphatase-labeled goat anti-human IgM, IgG, or IgA Ab (Southern Biotechnology Associates), diluted 1/1000 in sodium carbonate buffer, was used separately as secondary Ab in the replicate assays. Purified human IgM, IgG, or IgA (Biodesign International) was used as positive control, and culture supernatant from wells lacking B cells but containing all other medium additives was used as negative control.

We used an oligo-dT-based mRNA capture kit to isolate mRNA from the suspension generated by lysis of cells in the single B cell-derived clones (mRNA Capture kit; Roche Diagnostics). We used a single-tube strategy for RT-PCR amplification of the VH region (Titan One Tube RT-PCR System; Roche Diagnostics). The pooled PCR primer mixture that was designed to amplify Ab genes from all Ab gene families was described previously (22).

We ligated gel-extracted PCR products into a TA cloning vector (Promega), generated bacterial clones, and purified plasmid DNA from overnight bacterial cultures. Plasmid DNA was digested with restriction enzymes to identify clones with proper ligation. The nucleotide sequence of plasmid DNAs that contained a VH insert was determined using vector-specific primers and an automated DNA sequencer (Applied Biosystems).

We analyzed VH region sequences by comparison with the international ImMunoGeneTics (IMGT) information system database (http://imgt. cines.fr:8104) (23). The characteristics of individual sequences for VH regions are shown in the supplemental table.4 All sequences were submitted to GenBank (accession numbers EF177942 through EF178139). We characterized somatic mutations as replacement (R) or silent (S) mutations by comparing V region sequences with the corresponding IMGT germline sequences. The first 24 nucleotides of FR1 were encoded by PCR primers and therefore were not analyzed for mutations.

χ2 tests, with p values computed using Monte Carlo simulation with 500,000 replications, were used to compare proportions. Kruskal-Wallis tests were used to compare continuous variables (e.g., frequency of mutations, CDR3 lengths, numbers of insertions). All statistical analyses performed here assume independence between cells (i.e., ignore any correlation that might exist between measurements taken on different cells from the same individual).

All sorted RV VP6-specific naive (CD19+IgD+CD27, designated here virus-specific B cell subset 1 or VB1) and unclass-switched memory (CD19+IgD+CD27+, designated here VB2) or class-switched memory (CD19+IgDCD27+, designated here VB3) B cell samples exhibited a >95% purity in their naive or memory B cell phenotype after sorting. Data analysis was performed using FlowJo software (version 6.1 or above; Tree Star). Representative sorting data are shown in Fig. 1.

FIGURE 1.

Representative data from flow cytometric analyses demonstrating isolation of RV VP6-specific human naive and memory B cells with high purity. All cells plotted were gated before this analysis to be CD19+. All postsort samples contained >99% CD19+ B cells. Upper two panels, Presort and postsort samples are displayed for DLP-GFP binding; 1.1% of total B cells in the presort blood sample were positive for DLP-GFP; postsort DLP+IgD+CD27 (VB1), DLP+IgD+CD27+ (VB2), and DLP+IgDCD27+ (VB3) B cell samples were 61, 74, or 81% positive for DLP-GFP expression, respectively. Lower panel, The percentage of cells that were of naive (IgD+CD27) or memory (IgD+CD27+ or IgDCD27+) phenotypes are indicated by numbers in the quadrant gates. Postsort DLP+IgD+CD27 (VB1), DLP+IgD+CD27+ (VB2), and DLP+IgDCD27+ (VB3) B cell samples were shown to be 99, 96, or 96% pure for their phenotype, respectively.

FIGURE 1.

Representative data from flow cytometric analyses demonstrating isolation of RV VP6-specific human naive and memory B cells with high purity. All cells plotted were gated before this analysis to be CD19+. All postsort samples contained >99% CD19+ B cells. Upper two panels, Presort and postsort samples are displayed for DLP-GFP binding; 1.1% of total B cells in the presort blood sample were positive for DLP-GFP; postsort DLP+IgD+CD27 (VB1), DLP+IgD+CD27+ (VB2), and DLP+IgDCD27+ (VB3) B cell samples were 61, 74, or 81% positive for DLP-GFP expression, respectively. Lower panel, The percentage of cells that were of naive (IgD+CD27) or memory (IgD+CD27+ or IgDCD27+) phenotypes are indicated by numbers in the quadrant gates. Postsort DLP+IgD+CD27 (VB1), DLP+IgD+CD27+ (VB2), and DLP+IgDCD27+ (VB3) B cell samples were shown to be 99, 96, or 96% pure for their phenotype, respectively.

Close modal

Approximately 0.82% (range from 0.2 to 1.6%) of total circulating B cells isolated from blood samples of 10 healthy adults were shown to bind the RV VP6-specific GFP-DLPs by flow cytometric analysis. We examined the frequency of circulating RV VP6-specific or randomly selected naive or memory B cells in peripheral blood from these 10 healthy blood donors (Fig. 2). The mean frequency of RV VP6-specific B cells with the naive phenotype in these donors was 67.8% (range 29.2–85.4%). A mean of 22.6% (range 11.8–52.2%) of circulating RV VP6-specific B cells in our analysis were IgD+CD27+ memory B cells, and a mean of 7.7% (range 2.8–22.7%) of circulating RV VP6-specific B cells were IgDCD27+ memory B cells. The frequency of naive B cells was comparable to that in randomly selected B cells (mean 71.6%; range 57.4–89.4%) (p = 0.9). Interestingly, the frequency of unclass-switched IgD+CD27+ memory B cells was higher in RV VP6-specific B cells than that in randomly selected B cells (mean 13.6%; range 3.61–19.2%) (p = 0.013), while the frequency of IgDCD27+ memory B cells was lower (mean 13.9%; range 5.04–21.5%) (p = 0.045). The efficiency of isolation of Ag-specific cells, and the molecular cloning efficiency of variable gene isolation, is shown in the supplemental table.

FIGURE 2.

Higher frequency of unclass-switched memory cells in the RV-specific memory compartment than in randomly selected memory B cells. Frequencies of cells that were randomly selected or RV DLP-positive naive or memory phenotypes (IgD+CD27, ▵; IgD+CD27+, ⋄; IgDCD27+, □) from 10 independent samples are indicated on the y-axis. The mean values of each group were indicated as bars. On the x-axis, randomly selected IgD+CD27, IgD+CD27+, or IgDCD27+ B cells are referred to as CB1, CB2, or CB3 cells, respectively. DLP+IgD+CD27, DLP+IgD+CD27+, or DLP+IgDCD27+ B cells are designated VB1, VB2, or VB3 cells, respectively (**, p = 0.013; *, p = 0.045).

FIGURE 2.

Higher frequency of unclass-switched memory cells in the RV-specific memory compartment than in randomly selected memory B cells. Frequencies of cells that were randomly selected or RV DLP-positive naive or memory phenotypes (IgD+CD27, ▵; IgD+CD27+, ⋄; IgDCD27+, □) from 10 independent samples are indicated on the y-axis. The mean values of each group were indicated as bars. On the x-axis, randomly selected IgD+CD27, IgD+CD27+, or IgDCD27+ B cells are referred to as CB1, CB2, or CB3 cells, respectively. DLP+IgD+CD27, DLP+IgD+CD27+, or DLP+IgDCD27+ B cells are designated VB1, VB2, or VB3 cells, respectively (**, p = 0.013; *, p = 0.045).

Close modal

Isotypes of Ab produced by single B cell clones were determined for 38 RV-specific IgD+CD27 naive B cells (VB1), 46 RV-specific IgD+CD27+ memory B cells (VB2), and 80 RV-specific IgDCD27+ memory B cells (VB3) clones. Isotypes of Ab produced by randomly selected naive (designated here circulating B cell subset 1 or CB1) or IgD+ memory (designated here CB2) or IgD memory (designated here CB3) B cell clones from the same group of donors were reported previously and were used included for comparative purposes (24). As shown in Fig. 3, bottom panel, 94.7% of RV-specific IgD+CD27 naive B cells (VB1) and 93.5% of RV-specific IgD+CD27+ memory B cell clones (VB2) produced IgM. In contrast, the RV-specific IgDCD27+ memory B cell clones (VB3) produced IgM, IgG, or IgA. Interestingly, a higher frequency of IgM-producing B cell clones (55%) and a lower frequency of IgA-producing B cell clones (9%) were found in the RV-specific IgDCD27+ memory B cell group compared with the randomly selected IgDCD27+ memory B cell group, of which only 19% were IgM-producing clones (p < 0.0001), and 36% were IgA-producing clones (p < 0.0001). The frequencies of IgG-producing B cell clones were similar between RV specific and randomly selected IgDCD27+ memory B cells (39 vs 34%).

FIGURE 3.

Isotypes of Ab produced by RV VP6-specific memory B cells were more of IgM and less of IgA than that of randomly selected memory B cells. Culture supernatants from 38 clones of RV-specific IgD+CD27 naive B cells, 46 clones of RV-specific IgD+CD27+ memory B cells, and 80 clones of RV-specific IgDCD27+ memory B cells were used for determining isotypes of Abs produced following 21 days of culture. Approximately 94.7% of RV-specific IgD+CD27 naive B cell clones and 93.5% of RV-specific IgD+CD27+ memory B cell clones produced IgM. Higher frequency of IgM-producing B cell clones (55%) and lower frequency of IgA-producing B cell clones (9%) were found in the RV-specific IgDCD27+ memory B cell group compared with the randomly selected IgDCD27+ memory B cell group (p < 0.0001; p < 0.0001).

FIGURE 3.

Isotypes of Ab produced by RV VP6-specific memory B cells were more of IgM and less of IgA than that of randomly selected memory B cells. Culture supernatants from 38 clones of RV-specific IgD+CD27 naive B cells, 46 clones of RV-specific IgD+CD27+ memory B cells, and 80 clones of RV-specific IgDCD27+ memory B cells were used for determining isotypes of Abs produced following 21 days of culture. Approximately 94.7% of RV-specific IgD+CD27 naive B cell clones and 93.5% of RV-specific IgD+CD27+ memory B cell clones produced IgM. Higher frequency of IgM-producing B cell clones (55%) and lower frequency of IgA-producing B cell clones (9%) were found in the RV-specific IgDCD27+ memory B cell group compared with the randomly selected IgDCD27+ memory B cell group (p < 0.0001; p < 0.0001).

Close modal

As shown in Fig. 4, mutational analyses of VH genes isolated from RV-specific naive or memory single B cell clones demonstrated that comparable frequencies of mutations occurred in RV-specific naive (mean ± SD; 0.5 ± 0.8 bp) and IgD+CD27+ memory (11.0 ± 7.2 bp) B cells compared with those of randomly selected naive (1.2 ± 5.1 bp) and IgD+CD27+ memory (10.4 ± 7.5 bp) B cells. Interestingly, the frequency of somatic mutations in VH genes of RV-specific IgDCD27+ memory (14.1 ± 7.5 bp) B cells was significantly lower than frequency of somatic mutations in VH genes of randomly selected IgDCD27+ memory (20.3 ± 9.5 bp) (p < 0.0001) B cells. Further analysis demonstrated that mutations in all groups of RV-specific memory B cells were distributed across VH genes with an increased frequency of replacement mutations, especially in the CDR1, CDR2, and FR3 regions. Higher replacement/silent ratios (>3) were observed in the CDR regions of the RV-specific memory B cell groups than in the FR regions, despite the lower frequency of mutations in H chain variable genes of RV-specific IgDCD27+ memory B cells.

FIGURE 4.

Lower frequency of somatic mutations in RV VP6-specific IgDCD27+ memory B cells. Shown is the number of mutations identified in VH chain genes. A, RV VP6-specific circulating naive B cells (VB1). B, IgD+CD27+ memory B cells (VB2). C, IgDCD27+ memory B cells (VB3). For comparison, randomly selected circulating naive (CB1) or circulating memory B cells (CB2 or CB3) are shown. Mean values are indicated. VB1 cells possessed fewer mutations than did VB2 or VB3 cells (p < 0.0001). VB2 cells possessed fewer mutations than VB3 cells (p = 0.014). VB3 cells possessed fewer mutations than CB3 cells (p < 0.0001). Distributions of total somatic mutations and replacement/silent mutations were similar in all subsets of memory B cells. Higher replacement/silent ratios (>3) were detected in the CDR than FR regions in all memory cell groups.

FIGURE 4.

Lower frequency of somatic mutations in RV VP6-specific IgDCD27+ memory B cells. Shown is the number of mutations identified in VH chain genes. A, RV VP6-specific circulating naive B cells (VB1). B, IgD+CD27+ memory B cells (VB2). C, IgDCD27+ memory B cells (VB3). For comparison, randomly selected circulating naive (CB1) or circulating memory B cells (CB2 or CB3) are shown. Mean values are indicated. VB1 cells possessed fewer mutations than did VB2 or VB3 cells (p < 0.0001). VB2 cells possessed fewer mutations than VB3 cells (p = 0.014). VB3 cells possessed fewer mutations than CB3 cells (p < 0.0001). Distributions of total somatic mutations and replacement/silent mutations were similar in all subsets of memory B cells. Higher replacement/silent ratios (>3) were detected in the CDR than FR regions in all memory cell groups.

Close modal

We determined the nucleotide sequences of VH gene segments from 63 naive, 51 IgD+CD27+ memory, and 84 IgDCD27+ memory RV-specific B cell clones (Fig. 5). A significant difference was found in the distribution of VH gene family use between the three types of B cells (p = 0.04, χ2 test); 49% of VH gene segments in RV-specific IgD+CD27 naive B cells belonged to the VH1 family. The second and third most common VH families in RV-specific naive B cells were VH3 (28%) and VH4 (18%). In contrast, VH3 was the dominant VH family in RV-specific IgD+CD27+ and IgDCD27+ memory B cell clones (51 and 48%, respectively). VH1 segments were found in 26% of IgD+CD27+ and 29% of IgDCD27+ memory B cell clones. VH4 segments were found in 20% of IgD+CD27+ and 18% of IgDCD27+ memory B cell clones. Data about VH families in randomly selected naive and memory B cell clones from the same group of donors is shown for comparison. The VH3 dominance in RV-specific memory B cells was similar to the VH3 dominance in randomly selected naive and memory B cells (p > 0.5).

FIGURE 5.

RV VP6-specific B cell Ab repertoire is VH1 dominant in naive B cells, but evolves to be VH3 dominant in the two memory B cell groups. Nucleotide sequences of VH gene segments from 63 naive, 51 IgD+CD27+ memory, and 84 IgDCD27+ memory RV-specific B cell clones were determined. Frequencies of VH gene family use in all three RV-specific circulating naive and memory B cells are presented. Data from randomly selected naive or memory B cell clones from the same group of donors are shown for comparison. VH family gene usage was VH1 dominant in RV-specific naive B cells (p = 0.04), while VH family gene usage in RV-specific memory B cells was similar to the VH3 dominance in randomly selected naive and memory B cells (p > 0.5).

FIGURE 5.

RV VP6-specific B cell Ab repertoire is VH1 dominant in naive B cells, but evolves to be VH3 dominant in the two memory B cell groups. Nucleotide sequences of VH gene segments from 63 naive, 51 IgD+CD27+ memory, and 84 IgDCD27+ memory RV-specific B cell clones were determined. Frequencies of VH gene family use in all three RV-specific circulating naive and memory B cells are presented. Data from randomly selected naive or memory B cell clones from the same group of donors are shown for comparison. VH family gene usage was VH1 dominant in RV-specific naive B cells (p = 0.04), while VH family gene usage in RV-specific memory B cells was similar to the VH3 dominance in randomly selected naive and memory B cells (p > 0.5).

Close modal

D and JH family gene usages were analyzed in the three groups of RV-specific naive and memory B cells. Results from randomly selected naive and memory B cells in our previous report are shown for comparison. As shown in Fig. 6, striking similarities were found in the distributions of D and JH family genes. D3 was identified as the most common D family in all three groups, and was found in 40% of RV-specific IgD+CD27 naive B cells, 26% of RV-specific IgD+CD27+ memory B cells, and 46% of RV-specific IgDCD27+ memory B cell clones, respectively. JH4 was found to be the most commonly used JH family in all groups. 50% of RV-specific IgD+CD27 naive B cells, 40% of RV-specific IgD+CD27+ memory B cells, and 48% of RV-specific IgDCD27+ memory B cell clones used JH4 family genes. No difference was found in the distribution of D or JH gene family use between the three types of B cells (p > 0.3 for all comparisons, χ2 test).

FIGURE 6.

A, D3 family gene segment use dominated the repertoire of all subsets. Frequencies of D gene segments used in three RV-specific circulating naive and memory B cells are shown. Significant differences between cell types were not detected for D gene segment use (p > 0.3). B, Similar to randomly selected B cells, RV VP6-specific B cells use JH3 family genes as dominant JH genes. Frequencies of JH gene segments used in three RV-specific circulating naive and memory B cells are shown. No significant difference between cell types was found for JH gene segment use (p > 0.5).

FIGURE 6.

A, D3 family gene segment use dominated the repertoire of all subsets. Frequencies of D gene segments used in three RV-specific circulating naive and memory B cells are shown. Significant differences between cell types were not detected for D gene segment use (p > 0.3). B, Similar to randomly selected B cells, RV VP6-specific B cells use JH3 family genes as dominant JH genes. Frequencies of JH gene segments used in three RV-specific circulating naive and memory B cells are shown. No significant difference between cell types was found for JH gene segment use (p > 0.5).

Close modal

As shown in Fig. 7, a high frequency usage of the VH1–46 gene segment was noted in all three RV VP6-specific naive or memory B cell groups compared with VH1–46 gene segment usage in randomly selected B cell groups. There was a remarkably higher frequency of VH1–46 in the RV VP6-specific naive B cell group (28.6%) compared with the two RV VP6-specific memory B cell groups (7.8 and 8.3%, respectively) (p = 0.001). Interestingly, among a total of 198 VH chain genes isolated from single B cell clones, we identified a dominant VDJ recombination (VH1–46, D5–24, and JH4*02), which was identified in five clones (two were naive B cells, one was an IgD+CD27+ memory B cell, and two were IgDCD27+ memory B cells) isolated from four independent donors. The second most frequent VDJ recombination was of the VH1–46, D3–22, and JH4*02 gene segments, identified in four clones (all were naive B cells) isolated from three independent donors.

FIGURE 7.

VH1–46 is the dominant VH gene segment in RV VP6-specific naive B cells. Frequencies of VH gene segment use in all three RV-specific circulating naive or memory B cells are presented. Data from randomly selected naive or memory B cell clones from the same group of donors are shown for comparison. VH1–46 was the dominant VH gene segment used in RV-specific naive B cells, while the dominance became less apparent in memory cells (p = 0.001). In contrast, VH3–23 dominated the repertoire in all three subsets of randomly selected cells.

FIGURE 7.

VH1–46 is the dominant VH gene segment in RV VP6-specific naive B cells. Frequencies of VH gene segment use in all three RV-specific circulating naive or memory B cells are presented. Data from randomly selected naive or memory B cell clones from the same group of donors are shown for comparison. VH1–46 was the dominant VH gene segment used in RV-specific naive B cells, while the dominance became less apparent in memory cells (p = 0.001). In contrast, VH3–23 dominated the repertoire in all three subsets of randomly selected cells.

Close modal

Consistent with previous studies of randomly selected naive or memory B cells, analysis of H chain CDR3 regions suggested significantly shorter CDR3 lengths in RV-specific IgD+CD27+ memory B cells than in naive cells (24, 25). As shown in Fig. 8, the HCDR3 lengths for RV-specific IgD+CD27+ memory B cells were shorter than those of RV-specific IgD+CD27 naive B cells (p = 0.014) or those of RV-specific IgDCD27+ memory B cells (p = 0.001). The median HCDR3 length for RV-specific naive B cells was 45 bp (range from 27 to 72), whereas the median HCDR3 length for RV-specific IgD+CD27+ memory B cells was 39 bp (range from 21 to 78), and the median CDR3 length for RV-specific IgDCD27+ memory B cells was 48 bp (range from 21 to 72).

FIGURE 8.

Short VH chain CDR3 regions in RV VP6-specific IgD+CD27+ memory B cells. The HCDR3 lengths for three circulating RV-specific naive or memory B cells are shown with medians and 25th and 75th percentiles (interquartile range) indicated as bars. The HCDR3 lengths for RV-specific IgD+CD27+ memory B cells were shorter than those of RV-specific IgD+CD27 naive B cells (p = 0.014) and those of RV-specific IgDCD27+ memory B cells (p = 0.0012).

FIGURE 8.

Short VH chain CDR3 regions in RV VP6-specific IgD+CD27+ memory B cells. The HCDR3 lengths for three circulating RV-specific naive or memory B cells are shown with medians and 25th and 75th percentiles (interquartile range) indicated as bars. The HCDR3 lengths for RV-specific IgD+CD27+ memory B cells were shorter than those of RV-specific IgD+CD27 naive B cells (p = 0.014) and those of RV-specific IgDCD27+ memory B cells (p = 0.0012).

Close modal

We analyzed the VH-D and D-JH junctional sequences of RV-specific naive and memory B cells for evidence of the activity of insertional mechanisms. Comparison of these RV-specific sequences with those of randomly selected naive or memory B cells revealed that the characteristics of TdT-mediated alterations in the junctional sequences were similar. As indicated in Table I, the percentages of sequences that lacked N and P nucleotide additions at either VH-D or D-JH junctions were lower in the RV-specific IgDCD27+ memory B cells (6.3 and 16.4%) than those in RV-specific IgD+CD27+ memory B cells (14.6 and 25.0%), while RV-specific IgD+CD27 naive B cells demonstrated a higher percentage of sequences that lacked N and P nucleotide additions at D-JH junctions (26.7%) but a lower percentage of sequences that lacked N and P nucleotide additions at VH-D junctions (5.0%). Less N and P nucleotide additions were noted in both VH-D and D-JH junctions for IgD+CD27+ memory B cells than those in IgDCD27+ memory B cells (p < 0.001; p < 0.05), indicating that the shorter HCDR3 lengths in IgD+CD27+ memory B cells were a consequence of reduced numbers of N and P nucleotide insertions in both the VH-D and D-JH junctions. The total number of N and P nucleotides in RV-specific IgD+CD27 naive B cells and IgD+CD27+ memory B cells was significantly less than that in RV-specific IgDCD27+ memory B cells (p < 0.001).

Table I.

N and P nucleotide additions in the VH-D or D-JH junctions of the H chain CDR3 sequences of RV VP6-specific naive or memory B cells

SubsetIgDPercent Junctions Lacking N or P AdditionsMedian Number Nucleotide Additions at Indicated Junction (Interquartile Range)
VH-DD-JHVH-DD-JHCombined
Naive (CD27+ 5.0 26.7 3.0 (2.0, 6.0)a 3.0 (0.0, 7.0) 7.0 (5.0, 10.8)a 
Memory (CD27++ 14.6 25.0 4.0 (1.0, 6.0)a 2.5 (0.75, 6.0)b 8.0 (3.8, 11)a 
 − 6.3 16.4 7.0 (4.0, 10.0) 4.0 (1.0, 8.0) 12.0 (7.0, 17.5) 
SubsetIgDPercent Junctions Lacking N or P AdditionsMedian Number Nucleotide Additions at Indicated Junction (Interquartile Range)
VH-DD-JHVH-DD-JHCombined
Naive (CD27+ 5.0 26.7 3.0 (2.0, 6.0)a 3.0 (0.0, 7.0) 7.0 (5.0, 10.8)a 
Memory (CD27++ 14.6 25.0 4.0 (1.0, 6.0)a 2.5 (0.75, 6.0)b 8.0 (3.8, 11)a 
 − 6.3 16.4 7.0 (4.0, 10.0) 4.0 (1.0, 8.0) 12.0 (7.0, 17.5) 
a

Value of p < 0.001, compared to IgDCD27+ memory B cells.

b

Value of p < 0.05, compared to IgDCD27+ memory B cells.

Differing from our previous findings in randomly selected naive and memory B cells, examination of VH-D and D-JH junctions of the VH chain genes revealed striking similarities in the extent of exonuclease removal events between the VH, D, and JH gene ends in all B cell groups (24). As indicated in Table II, the number of exonuclease removals in the VH ends, JH ends, and 5′ and 3′ ends of the D region in RV-specific memory B cells were all similar to those in RV-specific naive B cells. However, further statistical analysis revealed significantly shorter D segment lengths in RV-specific IgD+CD27+ memory B cells compared with RV-specific naive B cells (p < 0.05), suggesting that the shorter HCDR3 lengths in RV-specific IgD+CD27+ memory B cells were also partly due to shorter D segment length in this group of B cells, especially when compared with RV-specific naive B cells.

Table II.

Nucleotide excision in the VH-D or D-JH junctions of the H chain CDR3 sequences of RV-specific naive or memory B cells

SubsetIgDMedian Number Nucleotide Deletions at Indicated Gene Segment End (Interquartile Range)Median D Length (Interquartile Range)
3′ of VH5′ of D3′ of D5′ of JH
Naive (CD27− 0 (0, 1.0) 2.5 (0, 6.0) 3.0 (0, 6.75) 4.0 (2.0, 6.0) 17.0 (14.0, 21.75) 
Memory (CD27+1.0 (0, 1.0) 4.0 (1.0, 7.0) 2.0 (1.0, 6.0) 5.5 (2.0, 8.0) 14.5 (12.0, 18.0)a 
 − 1.0 (0, 1.0) 3.0 (1.0, 6.5) 4.0 (1.0, 9.5) 3.0 (2.0, 8.0) 17.0 (13.5, 21.0) 
SubsetIgDMedian Number Nucleotide Deletions at Indicated Gene Segment End (Interquartile Range)Median D Length (Interquartile Range)
3′ of VH5′ of D3′ of D5′ of JH
Naive (CD27− 0 (0, 1.0) 2.5 (0, 6.0) 3.0 (0, 6.75) 4.0 (2.0, 6.0) 17.0 (14.0, 21.75) 
Memory (CD27+1.0 (0, 1.0) 4.0 (1.0, 7.0) 2.0 (1.0, 6.0) 5.5 (2.0, 8.0) 14.5 (12.0, 18.0)a 
 − 1.0 (0, 1.0) 3.0 (1.0, 6.5) 4.0 (1.0, 9.5) 3.0 (2.0, 8.0) 17.0 (13.5, 21.0) 
a

Value of p < 0.05 compared to naive cells.

These studies reveal molecular features of the unique VH1–46-immunodominant RV-specific human B cell repertoire, and the mechanisms by which the Ab repertoire can overcome a potentially limiting immunodominant primary response as cells are selected to progress into the two major memory B cell subsets. To our knowledge, this is the first study in which the molecular features of naturally occurring naive and memory Ag-specific B cells from humans have been performed in this way.

We found that the proportion of RV VP6-specific B cells that are memory cells represents a lower frequency and a less frequently isotype-switch phenotype compared with the proportion of randomly selected B cells that are memory cells. This high proportion of naive B cells in virus-specific clones in the circulation during infection, even in adults who likely have a history of many previous RV infections, is curious and unexplained (14, 15, 17). We showed previously a high frequency of naive B cells in the VP6 repertoire in adults, and a reduced frequency of somatic mutations in VP6-specific IgD memory B cells (15, 17). The Ab genes of RV-specific intestinal-homing memory (α4β7+ IgD) B cells in particular are almost completely devoid of somatic mutations, even though they are isotype switched (17). Our previous studies, however, considered IgD+-positive cells to be naive cells, whereas recent studies show that a third major subset of B cells exists within the unclass-switched IgD+ population (i.e., CD27+IgD+ cells) that are memory cells (18, 19). In the work here, we separately analyzed and compared the three major subjects of naive, unclass-switched memory, and class-switched memory cells from both RV-specific and randomly selected B cells. Detailed analysis of the molecular features of Ab genes from single cells revealed that the nature of the development of the RV-specific repertoire is distinct from that of the general repertoire. These findings are in agreement with more phenotypic studies in the literature that suggest RV VP6 is a unique Ag. For example, Parez et al. (14) reported that a large fraction of human naive B cells interacts with RV VP6 via surface Igs, and these VP6-reactive B cells were detected at similarly high frequencies in adult, infant, and neonatal samples. In mice, Blutt et al. (7) also showed robust activation of naive B cells, but not T cells, in Peyer’s patches and mesenteric lymph nodes during early RV infection. Using TCR gene deletion mouse models, these investigators further demonstrated that massive naive B cell activation occurred in a T cell-independent fashion (7). It should be noted, however, that this activation appears to be driven in the mouse model by VP7. Naive B cells also are the principal APCs associated with intestinal IgA production after s.c. RV injection in mice (8). A subset of B cells with optimal VP6-interacting germline-encoded surface Ig receptors in the naive B cell repertoire could explain the early involvement of naive B cells during RV infection.

We found that the proportion of RV VP6-specific B cells that are memory cells is low compared with all B cells, and these cells exhibit a less frequently isotype-switched phenotype. These unusual features of the RV repertoire may have important implications for understanding the mechanism underlying the relatively poor quality and duration of immunity to RV infection. Affinity maturation, enhanced by somatic mutations, is a critical factor in functional maturation of the human B cell response to RV (26). Also, secretory IgA is thought to be a significant effector molecule in RV immunity at the intestinal mucosal surface (3, 27, 28). Our studies here show, however, that the immunodominant human B cell response to VP6 specified by the VH1–46 gene segment is associated with reduced numbers of somatic mutations and reduced frequency of class switching, and thus likely reduced antiviral function.

The mechanism for this immunodominant response in the naive repertoire is not clear. Given the human RV studies showing a high proportion of naive B cells in the RV-specific circulating repertoire (14), the murine model evidence that a large portion of naive B cells expands in the absence of T cell help (7), and the observation of RV-protective Ab production in the TCR knockout mouse (29), it is intriguing to speculate that induction of RV-specific memory B cells occurs in the absence of efficient Th cell activity. Possibly the induction of RV-specific B cells occurs in a unique intestinal compartment independent of germinal centers, or in a unique cytokine milieu in the intestine. Alternatively, there may be a unique structural feature of the VP6 Ag that induces this distinct B cell pattern. Our studies do not resolve the mechanism underlying this phenomenon, but do show in great detail that the human B cell response to RV VP6 is unique, and suggest further studies of the induction of B cell immunity in the human intestine are needed.

One of the interesting aspects of the data here is the mechanism by which human antiviral Ab repertoires diversify following repeated infection, especially when a limited number of variable gene segments dominate the primary response. A limited number of variable gene segments is used in the primary response to RV, especially the VH1–46 gene segment that encodes the BCR in over a quarter of all VP6-specific cells, even though this segment represents <4% of the random repertoire. The dominance of this VH gene segment appears to stem from the fact that the HCDR1 and HCDR2 loops specified by VH1–46 exhibit an optimal Ab-combining site for loops on the surface of VP6 with a high level of shape complementarity and resulting optimal affinity for a primary response (26). Our original report on RV-specific B cell repertoire also suggested VH1–46 gene usage in B cells sorted to be VP7 specific (15). Our subsequent studies, however, suggested that these clones were VP6 specific and likely sorted by the DLPs displaying VP6 that resulted from the loss of VP7 on the relatively unstable triple-layered particles used for sorting. One would expect that during a secondary immune response to RV, the ∼30% of RV-specific cells in the resulting primary repertoire that use surface receptors specified by VH1–46 would affinity-mature through somatic hypermutation and would become even more dominant in the repertoire. We expected that the VP6-specific IgD+ and IgD memory cells would exhibit a progressively more oligoclonal repertoire, fully dominated by VH1–46 Ab-expressing cells. Unexpectedly, we found the opposite to be true, because the combinatorial diversity of the memory cell repertoires increased in breadth. VH1–46 was present in all three RV VP6-specific circulating human naive and memory B cell subsets; however, the data clearly show increased VH diversity in the two memory B cell repertoires. The frequency of use of VH1–46 decreased as the profile of memory cell gene segments began to resemble the more broad VH3-dominated repertoire of randomly selected B cells. Our detailed analysis demonstrated that memory RV VP6-specific B cells diversify their Ab repertoire by means of expansion of clones with broader use of VH gene segments, increased junctional diversity, and accumulation of somatic mutations. We propose a model for diversification in which clones with subdominant VH, D, and JH segments are present in the primary responding repertoire but are less frequently identified because of reduced affinity of these Abs compared with the germline-encoded VH1–46 Abs. Following repeated exposure to Ag during secondary responses, low-affinity receptors specified by subdominant gene segments likely are able to “catch up” in terms of affinity through somatic hypermutation, and expand to populate the repertoire in increased proportions. Why the increased precursor frequency of VH1–46 cells in the naive repertoire does not allow these clones to increase their dominance is not clear. VH3 and other gene families that are underrepresented in the naive RV-specific repertoire may have an advantage due to non-Ag-specific factors that are not understood.

It is not clear why both the RV-specific and the general B cell repertoire are driven toward a VH3-dominant profile in memory cells, although it is clear that this transition does occur. We have shown previously that the frequency of use of particular VH, D, JH, VL, and JL gene segments in the general repertoire is exceedingly highly regulated, is always VH3 dominant, and is virtually identical in naive, and IgD+CD27+ and IgDCD27+ memory cells of all individuals tested (24). Even taking into account that the VH3 gene family contains more segments than other families, the circulating repertoire exhibits a strong VH3 bias.

Finally, the large number of VP6-specific clones that we isolated allowed us to identify apparently preferred VH-D-JH combinations for Abs binding to the VP6 protein. Given the large number of potential combinations of variable gene segments that exist, one would not expect to randomly identify particular combinations. We found a dominant VDJ recombination (VH1–46, D5–24, and JH4*02), and the second most frequent VDJ recombination (VH1–46, D3–22, and JH4*02). These combinations likely specify H chain protein sequences contributing to optimal Ag-combining sites for VP6. Interestingly too, the identification of these clones suggests that the HCDR3 region contributes to optimization of interactions of VH1–46 Abs with VP6. The immunodominance of VH1–46 even in combination with diverse DH and J segments implies that the principal interacting surface of the VH1–46 Abs is initially the HCDR2 loop, which is specified in the VH gene segment sequence independent of DH and J. The VH-D junction, D segment, and D-JH junction specify the HCDR3 loop, which likely optimizes the interaction in the several preferred VDJ combinations for these Abs. This HCDR2-dominated Ag-Ab interaction is somewhat unusual, because for most Ag-Ab interactions, the HCDR3 is the principal interacting structural element of the Ab (30, 31, 32, 33, 34).

We thank DNAX for use of the CD154-expressing cell line, Dr. Rudolf H. Zubler for the EL4-B5 cell line, and Dr. Harry Greenberg for RV reagents and advice.

The authors have no financial conflict of interest.

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 a grant from the National Institute of Allergy and Infectious Diseases (NIAID; R01 AI-57933). C.T. was a fellow of the NIAID Molecular Basis in Infectious Diseases Training Program (T32 AI-07474). The Vanderbilt Flow Cytometry Core Laboratory was supported by the Vanderbilt Ingram Cancer Center (P30 CA68485). Clinical support was provided by the Vanderbilt General Clinical Research Center (Grant M01 RR-00095 National Center for Research Resources, National Institutes of Health).

3

Abbreviations used in this paper: RV, rotavirus; VLP, virus-like particle; DLP, double-layered particle; VB, virus-specific B cell subset; CB, circulating (randomly selected) B cell subset; FR, framework region.

4

The online version of this article contains supplemental material.

1
Parashar, U. D., C. J. Gibson, J. S. Bresse, R. I. Glass.
2006
. Rotavirus and severe childhood diarrhea.
Emerg. Infect. Dis.
12
:
304
-306.
2
Bishop, R. F., G. P. Davidson, I. H. Holmes, B. J. Ruck.
1973
. Virus particles in epithelial cells of duodenal mucosa from children with acute non-bacterial gastroenteritis.
Lancet
2
:
1281
-1283.
3
Burns, J. W., A. A. Krishnaney, P. T. Vo, R. V. Rouse, L. J. Anderson, H. B. Greenberg.
1995
. Analyses of homologous rotavirus infection in the mouse model.
Virology
207
:
143
-153.
4
Franco, M. A., H. B. Greenberg.
1995
. Role of B cells and cytotoxic T lymphocytes in clearance of and immunity to rotavirus infection in mice.
J. Virol.
69
:
7800
-7806.
5
Dharakul, T., L. Rott, H. B. Greenberg.
1990
. Recovery from chronic rotavirus infection in mice with severe combined immunodeficiency: virus clearance mediated by adoptive transfer of immune CD8+ T lymphocytes.
J. Virol.
64
:
4375
-4382.
6
VanCott, J. L., M. M. McNeal, J. Flint, S. A. Bailey, A. H. Choi, R. L. Ward.
2001
. Role for T cell-independent B cell activity in the resolution of primary rotavirus infection in mice.
Eur. J. Immunol.
31
:
3380
-3387.
7
Blutt, S. E., K. L. Warfield, D. E. Lewis, M. E. Conner.
2002
. Early response to rotavirus infection involves massive B cell activation.
J. Immunol.
168
:
5716
-5721.
8
Coffin, S. E., S. L. Clark, N. A. Bos, J. O. Brubaker, P. A. Offit.
1999
. Migration of antigen-presenting B cells from peripheral to mucosal lymphoid tissues may induce intestinal antigen-specific IgA following parenteral immunization.
J. Immunol.
163
:
3064
-3070.
9
Keating, G. M..
2006
. Rotavirus vaccine (RotaTeq).
Paediatr. Drugs
8
:
197
-202.
10
Offit, P. A., H. F. Clark.
2006
. RotaTeq: a pentavalent bovine–human reassortant rotavirus vaccine.
Pediatr. Ann.
35
:
29
-34.
11
Estes, M. K., A. Z. Kapikian.
2006
. Rotaviruses. D.M. Knipe, and P. M. Howley, eds.
Fields Virology
5th Ed.
1917
-1974. Philadelphia, Lippincott Williams & Wilkins.
12
Burns, J. W., M. Siadat-Pajouh, A. A. Krishnaney, H. B. Greenberg.
1996
. Protective effect of rotavirus VP6-specific IgA monoclonal antibodies that lack burneutralizing activity.
Science
272
:
104
-107.
13
Schwartz-Cornil, I., Y. Benureau, H. Greenberg, B. A. Hendrickson, J. Cohen.
2002
. Heterologous protection induced by the inner capsid proteins of rotavirus requires transcytosis of mucosal immunoglobulins.
J. Virol.
76
:
8110
-8117.
14
Parez, N., A. Garbarg-Chenon, C. Fourgeux, F. Le Deist, A. Servant-Delmas, A. Charpilienne, J. Cohen, I. Schwartz-Cornil.
2004
. The VP6 protein of rotavirus interacts with a large fraction of human naive B cells via surface immunoglobulins.
J. Virol.
78
:
12489
-12496.
15
Weitkamp, J. H., N. Kallewaard, K. Kusuhara, E. Bures, J. V. Williams, B. LaFleur, H. B. Greenberg, J. E. Crowe, Jr.
2003
. Infant and adult human B cell responses to rotavirus share common immunodominant variable gene repertoires.
J. Immunol.
171
:
4680
-4688.
16
Weitkamp, J. H., B. J. Lafleur, J. E. Crowe, Jr.
2006
. Rotavirus-specific CD5+ B cells in young children exhibit a distinct antibody repertoire compared with CD5 B cells.
Hum. Immunol.
67
:
33
-42.
17
Weitkamp, J. H., N. L. Kallewaard, A. L. Bowen, B. J. Lafleur, H. B. Greenberg, J. E. Crowe, Jr.
2005
. VH1–46 is the dominant immunoglobulin heavy chain gene segment in rotavirus-specific memory B cells expressing the intestinal homing receptor α4β7.
J. Immunol.
174
:
3454
-3460.
18
Klein, U., K. Rajewsky, R. Kuppers.
1998
. 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
:
1679
-1689.
19
Shi, Y., K. Agematsu, H. D. Ochs, K. Sugane.
2003
. Functional analysis of human memory B-cell subpopulations: IgD+CD27+ B cells are crucial in secondary immune response by producing high affinity IgM.
Clin. Immunol.
108
:
128
-137.
20
Weller, S., M. C. Braun, B. K. Tan, A. Rosenwald, C. Cordier, M. E. Conley, A. Plebani, D. S. Kumararatne, D. Bonnet, O. Tournilhac, et al
2004
. Human blood IgM “memory” B cells are circulating splenic marginal zone B cells harboring a prediversified immunoglobulin repertoire.
Blood
104
:
3647
-3654.
21
Kikuchi, K., Z. X. Lian, G. X. Yang, A. A. Ansari, S. Ikehara, M. Kaplan, H. Miyakawa, R. L. Coppel, M. E. Gershwin.
2005
. Bacterial CpG induces hyper-IgM production in CD27+ memory B cells in primary biliary cirrhosis.
Gastroenterology
128
:
304
-312.
22
Weitkamp, J. H., N. Kallewaard, K. Kusuhara, D. Feigelstock, N. Feng, H. B. Greenberg, J. E. Crowe, Jr.
2003
. Generation of recombinant human monoclonal antibodies to rotavirus from single antigen-specific B cells selected with fluorescent virus-like particles.
J. Immunol. Methods
275
:
223
-237.
23
Ruiz, M., V. Giudicelli, C. Ginestoux, P. Stoehr, J. Robinson, J. Bodmer, S. G. Marsh, R. Bontrop, M. Lemaitre, G. Lefranc, et al
2000
. IMGT, the international ImMunoGeneTics database.
Nucleic Acids Res.
28
:
219
-221.
24
Tian, C., G. K. Luskin, K. M. Dischert, J. N. Higginbotham, B. E. Shepherd, J. E. Crowe, Jr.
2007
. Evidence for preferential Ig gene usage and differential TdT and exonuclease activities in human naive and memory B cells.
Mol. Immunol.
44
:
2173
-2183.
25
Tsuiji, M., S. Yurasov, K. Velinzon, S. Thomas, M. C. Nussenzweig, H. Wardemann.
2006
. A checkpoint for autoreactivity in human IgM+ memory B cell development.
J. Exp. Med.
203
:
393
-400.
26
Kallewaard, N. L., B. A. McKinney, Y. Gu, A. Chen, B. V. V. Prasad, and J. E. Crowe, Jr. Functional maturation of the human antibody response to rotavirus. J. Immunol. In press.
27
Williams, M. B., J. R. Rose, L. S. Rott, M. A. Franco, H. B. Greenberg, E. C. Butcher.
1998
. The memory B cell subset responsible for the secretory IgA response and protective humoral immunity to rotavirus expresses the intestinal homing receptor, α4β7.
J. Immunol.
161
:
4227
-4235.
28
Feng, N., J. W. Burns, L. Bracy, H. B. Greenberg.
1994
. Comparison of mucosal and systemic humoral immune responses and subsequent protection in mice orally inoculated with a homologous or a heterologous rotavirus.
J. Virol.
68
:
7766
-7773.
29
Franco, M. A., H. B. Greenberg.
1997
. Immunity to rotavirus in T cell deficient mice.
Virology
338
:
169
-179.
30
Martin, T., S. F. Duffy, D. A. Carson, T. J. Kipps.
1992
. Evidence for somatic selection of natural autoantibodies.
J. Exp. Med.
175
:
983
-991.
31
Martin, T., R. Crouzier, J. C. Weber, T. J. Kipps, J. L. Pasquali.
1994
. Structure-function studies on a polyreactive (natural) autoantibody: polyreactivity is dependent on somatically generated sequences in the third complementarity-determining region of the antibody heavy chain.
J. Immunol.
152
:
5988
-5996.
32
Chen, C., M. P. Stenzel-Poore, M. B. Rittenberg.
1991
. Natural auto- and polyreactive antibodies differing from antigen-induced antibodies in the H chain CDR3.
J. Immunol.
147
:
2359
-2367.
33
Ichiyoshi, Y., P. Casali.
1994
. Analysis of the structural correlates for antibody polyreactivity by multiple reassortments of chimeric human immunoglobulin heavy and light chain V segments.
J. Exp. Med.
180
:
885
-895.
34
Ditzel, H. J., K. Itoh, D. R. Burton.
1996
. Determinants of polyreactivity in a large panel of recombinant human antibodies from HIV-1 infection.
J. Immunol.
157
:
739
-749.

Supplementary data