Memory B cells expressing the intestinal homing marker α4β7 are important for protective immunity against human rotavirus (RV). It is not known whether the B cell repertoire of intestinal homing B cells differs from B cells of the systemic compartment. In this study, we analyzed the RV-specific VH and VL repertoire in human IgD B cells expressing the intestinal homing marker α4β7. The mean frequency of RV-specific B cells in the systemic compartment of healthy adult subjects was 0.6% (range, 0.2–1.2). The mean frequency of IgD B cells that were both RV specific and α4β7 was 0.04% (range, 0.01–0.1), and a mean of 10% (range, 1–32) of RV-specific peripheral blood human B cells exhibited an intestinal homing phenotype. We previously demonstrated that VH1–46 is the dominant Ab H chain gene segment in RV-specific systemic B cells from adults and infants. RV-specific systemic IgD or intestinal homing IgD4β7+ B cells in the current study also used the gene segment VH1–46 at a high frequency, while randomly selected B cells with those phenotypes did not. These data show that VH1–46 is the immunodominant gene segment in human RV-specific effector B cells in both the systemic compartment and in intestinal homing lymphocytes. The mean replacement/silent mutation ratio of systemic compartment IgD B cells was >2, consistent with a memory phenotype and antigenic selection. Interestingly, RV-specific intestinal homing IgD4β7+ B cells using the VH1–46 gene segment were not mutated, in contrast to systemic RV-specific IgD B cells.

Human rotaviruses (RV)3 replicate in the small intestine and represent the most common cause of infectious gastroenteritis in children. Previous data in animal studies suggested that a principal component of protective immunity for RV resides in intestinal homing B cells, recognized as IgD B cells expressing the intestinal homing receptor α4β7 (1). It is not known whether the B cell repertoire of cells that traffic to the intestine differs from those of the systemic compartment. We previously demonstrated that VH1–46 is the dominant Ab H chain gene segment in RV-specific systemic compartment human B cells from adults and infants. In the current study, we analyzed the RV-specific VH and VL repertoire in human IgD B cells expressing the intestinal homing marker α4β7 to determine whether the systemic and intestinal homing populations utilized common or distinct repertoires.

Lymphoid cells recirculate continuously from the systemic compartment to efferent lymph through secondary lymphoid tissues, to which they arrive by traversing the cuboidal endothelium of high endothelial venules. Integrin interactions are essential for the proper trafficking of B cells exposed to Ags in the intestine, to Peyer’s patches, to inductive lymphoid tissues, and finally to the lamina propria effector site. The intestinal homing pattern of GALT-derived effector B cells was suggested in previous repertoire studies in which the Ig VH region genes from human Peyer’s patch germinal center B cells were found to be clonally related to those of ileal lamina propria plasma cells (2). Homing of Ag-primed GALT-derived B and T cells to the human intestinal mucosa is determined mainly by their expression of high levels of the α4β7 integrin in the absence of L-selectin expression as well as coexpression of selected chemokine receptors, such as CCR9 (3). Systemic IgD+ B cells homogeneously express high levels of the α4β7 molecule, but IgD memory B cells comprise distinct α4β7+ and α4β7 subsets (4). Naive and memory α4β7+ B cells can bind the mucosal addressin cell adhesion molecule 1.

RV replicates in the intestinal epithelium, and it is likely that RV immunity in humans is mediated by effector cells in the gut. We previously demonstrated that protective intestinal anti-RV B cell immunity in the mouse depends on α4β7 integrin expression (1). We also showed that adoptively transferred IgD4β7+ B cells, but not IgD+4β7+ or IgD4β7 B cells, cleared chronic RV infection in RAG-2-deficient mice that lack T and B cells (5). Recent studies determined that human RV-specific surface Ig+ B cells often express the α4β7 molecule, as detected by a novel mAb specific for that heterodimer (6). We reasoned that analysis of the Ab repertoire in human RV-specific B cells should focus on these intestinal homing cells.

Our previous studies demonstrated that VH1–46 is the dominant Ab H chain gene segment in RV-specific systemic B cells from both adults and infants (7, 8). We showed that recombinant Abs specified by immunodominant Ab gene segments of RV-specific B cells, including VH1–46, bind to RV particles or RV-infected cells. RV-specific B cells that use the VH1–46 gene segment are the most commonly isolated virus-specific B cells in the systemic compartment. These findings allowed us to test in the current study whether the B cell repertoire of the systemic compartment is independent of, or concordant with, the repertoire of intestinal homing B cells directed to a common and important intestinal viral pathogen in humans. The goal for the current study was to isolate RV-specific memory B cells that exhibited an intestinal homing phenotype (expression of intestinal homing receptor α4β7 in IgD B cells) and to compare the Ab repertoire of these effector cells with that of systemic RV-specific IgD or randomly selected intestinal homing B cells. We found that randomly selected intestinal homing B cells exhibit a different repertoire bias compared with randomly selected B cells from the systemic compartment. In contrast, systemic and intestinal homing RV-specific B cells share the overrepresentation of the VH1–46 gene segment, suggesting that the systemic and mucosal compartments share a common RV-specific repertoire.

We isolated single human RV VP6-specific B cell clones from peripheral blood samples using a single-cell flow cytometry sorting method, as described previously (8). Ag-specific B cells were identified by staining with GFP-labeled virus-like particles (VLP) (8). For this study, we used GFP-labeled RV VP2/6 double-layered particles (DLP). GFP-VLP for single-cell sorting were produced by coinfection of Sf9 insect cells with recombinant baculoviruses as described elsewhere (1). 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 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.

We obtained lymphocytes from blood from two different types of donors. 1) White blood cells were obtained from individual leukocyte reduction filters from 12 healthy Red Cross blood donors as described previously (9). For analysis of RV-specific α4β7-expressing IgD B cells, white blood cells from leukocyte reduction filters from seven separate donors were collected. For randomly selected (not Ag-specific) α4β7-expressing IgD B cell controls, we sampled blood cells from leukocyte reduction filters from three separate donors. 2) Blood for analysis of RV-specific α4β7-expressing IgD B cells was also obtained from one healthy adult donor by venipuncture. Four of 37 clones analyzed for VH gene segment usage with this phenotype were from this donor. We isolated PBMC by density gradient centrifugation on lymphocyte separation medium and separated CD19+ cells with paramagnetic beads according to the manufacturer’s instructions (Dynal). All B cells were obtained by the same investigator in the same laboratory during the same time period by an identical procedure. We stained magnetically isolated CD19+ cells with GFP-labeled RV DLP and anti-IgD-PerCP as described elsewhere (7). In addition, we included a staining with 2 μl per 106 cells of an anti-α4β7-specific mAb conjugated to PE (50 μg/ml, kindly provided by E. Butcher, Stanford University, Stanford, CA). Commercially available isotype-matched Abs were used to control for nonspecific binding. We performed flow cytometric analysis and sorting using a FACStarPlus cytometer (BD Biosciences). Single RV DLP+/IgD4β7+ or single RV DLP+/IgD4β7unselected B cells were collected as one cell per well into 96-well cell culture plates using the automated single-cell deposition unit of the cytometer. Controls were single RV DLP+/IgDunselected4β7unselected and single RV DLP+/IgD4β7unselected B cells. These two populations were analyzed previously to define the human RV-specific B cell repertoire in adults and are included here for comparative purposes (7). We refer to α4β7unselected B cells from these studies as “systemic” since previous work has shown that the vast majority of peripheral blood B cells are not IgD4β7+ intestinal homing cells.

For the expansion of single B cells into clones, we used a culture system as recently described (8). 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. Secreted total human Ig or RV-specific Abs were detected by ELISA using RV DLP as Ag to confirm the specificity of the B cell clone for RV before RT-PCR for recovery of Ab genes (8). The DLP used in this ELISA differed from those used in sorting, in that they were naturally occurring DLP isolated from strain rhesus rotavirus-infected MA104 cells. The rationale for using this Ag in the ELISA screening was to avoid detection of clones secreting Abs to GFP, if such clones were inadvertently selected in the flow cytometry procedure. We analyzed only clones that were determined to have RV specificity by at least two different methods in that 1) they bound to RV Ags by flow cytometric analysis and 2) the Abs secreted by the resulting B cell clone showed specificity of binding by strict criteria in the Ag-specific ELISA. To further confirm that the criteria for selection of RV-specific cells was effective, we generated recombinant Abs containing the immunodominant VH gene segment 1–46 and showed that they bind to RV-infected cells or purified RV particles (7). We found these purified Fabs bound in a dose-dependent and specific manner to DLP, but not to complete rhesus rotavirus particles, showing again their specificity for VP6.

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 secreting human Ig and RV-specific Abs. We used a one-tube system for RT-PCR amplification of VH and VL regions (Titan One Tube RT-PCR System; Roche Diagnostics). The pooled PCR primer mixture that we used was designed to amplify Ab genes from all Ab gene families and was recently described (8).

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 or a VL insert was determined using vector-specific primers and an automated DNA sequencer (Applied Biosystems).

We analyzed VH or VL region sequences by comparison to the International ImMunoGeneTics Information System (http://imgt.cines.fr). All sequences were submitted to GenBank (accession numbers AY686868 through AY686939).

Fisher’s exact tests were used to examine overall differences in proportions of frequencies of use of VH, D, and JH gene family members, VH gene segments, and Vκ, Jκ, Vλ, and Jλ gene family members in the different B cell subsets that were selected.

We examined the frequency of RV DLP-binding IgD or IgD4β7+ B cells in peripheral blood from 11 healthy blood donors (Table I). The mean frequency of RV-specific B cells in these donors was 0.6% (range, 0.2–1.2%). A mean of 25% (range, 1–49%) of RV-specific B cells in our analysis lacked IgD expression, suggesting that they were memory B cells. Of those, a mean of 57% (range, 2–100%) also expressed the α4β7 marker. Within the total B cell population, the frequency of IgD B cells that were RV specific and expressed the intestinal homing receptor α4β7 was 0.04% (range, 0.01–0.1%). Therefore, a mean of 10% (range, 1–32%) of RV-specific B cells in the peripheral blood of previously infected healthy adults were also exhibiting an intestinal homing phenotype.

Table I.

Distribution of surface markers in B cells obtained from previously RV-infected healthy adultsa

DonorNo. of B Cells Sorted (×106)% Total CD19+ B Cells That Exhibit the Indicated Phenotype% RV-Specific (DLP-GFP+) B Cells That also Exhibited the Indicated Phenotype
RV VP6 specificity (DLP-GFP+)RV specificity and intestinal homing (DLP-GFP+/IgD4β7+)Intestinal homing (IgD4β7+)
4.5 1.2 0.01 
0.7 0.01 
3.5 0.7 0.02 
0.6 0.02 
0.9 0.01 
0.6 0.02 
0.5 0.4 0.10 27 
0.3 0.10 32 
0.5 0.07 15 
10 30 0.2 0.03 15 
11 0.2 0.02 10 
Mean (range) 5.8 (0.5–30) 0.6 (0.2–1.2) 0.04 (0.01–0.1) 10 (1–32) 
DonorNo. of B Cells Sorted (×106)% Total CD19+ B Cells That Exhibit the Indicated Phenotype% RV-Specific (DLP-GFP+) B Cells That also Exhibited the Indicated Phenotype
RV VP6 specificity (DLP-GFP+)RV specificity and intestinal homing (DLP-GFP+/IgD4β7+)Intestinal homing (IgD4β7+)
4.5 1.2 0.01 
0.7 0.01 
3.5 0.7 0.02 
0.6 0.02 
0.9 0.01 
0.6 0.02 
0.5 0.4 0.10 27 
0.3 0.10 32 
0.5 0.07 15 
10 30 0.2 0.03 15 
11 0.2 0.02 10 
Mean (range) 5.8 (0.5–30) 0.6 (0.2–1.2) 0.04 (0.01–0.1) 10 (1–32) 
a

PBMCs were isolated by density centrifugation and CD19+ cells were separated with paramagnetic beads. This procedure yielded ≥99% CD19+ viable cells when analyzed by flow cytometry. These cells were subsequently stained as follows: for RV VP6 with GFP-labeled DLPs, for IgD B cells with biotinylated goat anti-human IgD mAbs and streptavidin PerCP, and for α4β7 with an anti-α4β7-specific mAb conjugated to PE. Commercially available isotype-matched Abs were used to control for nonspecific binding. The percentages in column 4 were calculated from columns 3 and 5.

We analyzed 18 VH segments from six separate donors to characterize the Ab repertoire in randomly selected intestinal homing (IgD4β7+) B cells. We previously expanded randomly selected CD19+ B cells (from 3 donors) and RV-specific CD19+memory B cells (from 2 donors) and determined the nucleotide sequences of VH segments from 83 or 18 B cells from those groups, respectively (GenBank accession numbers AF453004 through AF453079, AF453218 through AF453222, and AF452986 through AF453003) (7). These data are included here for comparative purposes. Forty-nine percent of VH segments in randomly selected B cells belonged to the VH3 family (Fig. 1). The second most common VH family in randomly selected B cells was VH4, found in 20% of cells. In comparison, randomly selected IgD4β7+ B cells exhibited a trend toward stronger VH4 bias (44% of all clones) when compared with randomly selected peripheral blood B cells. However, this difference was not statistically significant (p = 0.13).

FIGURE 1.

Frequency of VH family utilization in randomly selected systemic (▪) or randomly selected IgD4β7+ intestinal homing B cells (□) compared with clones from RV-specific systemic IgD (▦) or RV-specific IgD4β7+ intestinal homing B cells (▨). Single RV-specific B cells from healthy blood donors were sorted, amplified in culture, and tested for specificity by ELISA followed by sequencing of expressed VH gene segments. Eighty-three clones were analyzed for randomly selected systemic cells, 18 clones for randomly selected IgD4β7+ intestinal homing cells, 18 clones for RV-specific systemic IgD cells, and 33 clones for RV-specific IgD4β7+ intestinal homing B cells. The 83 individual VH sequences from randomly selected systemic B cells and 18 individual VH sequences from RV-specific systemic IgD are published as supplemental data in Ref. 7 . Comparisons in proportions for this figure were performed by using the Fisher’s exact test statistic: comparison between randomly selected peripheral B cells and randomly selected intestinal homing B cells, p = 0.133. Comparison between randomly selected intestinal homing B cells and RV-specific intestinal homing B cells, p = 0.089.

FIGURE 1.

Frequency of VH family utilization in randomly selected systemic (▪) or randomly selected IgD4β7+ intestinal homing B cells (□) compared with clones from RV-specific systemic IgD (▦) or RV-specific IgD4β7+ intestinal homing B cells (▨). Single RV-specific B cells from healthy blood donors were sorted, amplified in culture, and tested for specificity by ELISA followed by sequencing of expressed VH gene segments. Eighty-three clones were analyzed for randomly selected systemic cells, 18 clones for randomly selected IgD4β7+ intestinal homing cells, 18 clones for RV-specific systemic IgD cells, and 33 clones for RV-specific IgD4β7+ intestinal homing B cells. The 83 individual VH sequences from randomly selected systemic B cells and 18 individual VH sequences from RV-specific systemic IgD are published as supplemental data in Ref. 7 . Comparisons in proportions for this figure were performed by using the Fisher’s exact test statistic: comparison between randomly selected peripheral B cells and randomly selected intestinal homing B cells, p = 0.133. Comparison between randomly selected intestinal homing B cells and RV-specific intestinal homing B cells, p = 0.089.

Close modal

In previous work, we observed a strong VH1/VH4 gene family bias in RV-specific B cells in the blood and a marked VH4 family bias for RV VP6-specific B cells, with a VH4 family frequency of 71% (7). These data suggested that VH gene segments from these two gene families specify Abs that are inherently fit for RV binding. In the present studies, we also observed a VH1/VH4 gene family bias in RV-specific systemic IgD B cells and found this Ag selection bias also characterized RV-specific IgD4β7+ B cells (Fig. 1). Eighteen clones total from 2 donors were analyzed for RV-specific systemic IgD cells, and 33 clones from 8 donors were analyzed for RV-specific IgD4β7+ intestinal homing B cells. Both RV-specific IgD α4β7-unselected and IgD intestinal homing B cells exhibited a strong VH1 gene family bias, with 61 or 45% of all clones belonging to the VH1 family, respectively. However, since randomly selected intestinal homing memory B cells also exhibited a VH4 family bias, the overall difference in VH family usage between RV-specific and randomly selected B cells in the intestinal homing compartment was not statistically significant (p = 0.09).

As shown by us and others, the most commonly used VH gene segment in randomly selected systemic compartment B cells is the VH3–23 gene segment. Interestingly, we observed a statistically significant difference in VH segment usage between randomly selected peripheral B cells and randomly selected intestinal homing B cells (Fig. 2). In contrast to randomly selected peripheral B cells in which 10% of all clones (8 of 83) were found to be using the VH3–23 gene segment, only 6% (1 of 18) of randomly selected intestinal homing B cells used this segment, a 40% reduction. By comparison, a larger proportion of intestinal homing B cells (16% or 3 of 18) used the VH4–39 gene segment (p = 0.03 for comparison of distribution of gene segments).

FIGURE 2.

Dominant VH gene segment utilization in RV-specific human B cell clones expressing the intestinal homing receptor α4β7+ compared with that of randomly selected B cells. The relative frequency of three dominant VH gene segments in RV-specific intestinal homing B cells (▨) is plotted in comparison to the frequency of use of these VH gene segments in randomly selected (▪) or RV-specific (▦) systemic IgD B cells or randomly selected intestinal homing B cells (□). The frequency of use of the VH3–23 segment (the dominant segment in randomly selected systemic blood B cells) is shown for comparative purposes. Eighty-three clones were analyzed for randomly selected systemic cells, 18 for randomly selected IgD4β7+ intestinal homing cells, 18 for RV-specific systemic IgD cells, and 33 for RV-specific IgD4β7+ intestinal homing B cells. The 83 individual VH sequences from randomly selected systemic B cells and 18 individual VH sequences from RV-specific systemic IgD are published as supplemental data in Ref. 7 . Comparisons in proportions for this figure were performed by using the Fisher’s exact test statistic. When proportions of VH gene segment usages were compared between randomly selected peripheral B cells and randomly selected intestinal homing B cells using this method, the p value was 0.032. Comparison of proportions between randomly selected intestinal homing B cells and RV-specific intestinal homing B cells resulted in a p value of 0.002.

FIGURE 2.

Dominant VH gene segment utilization in RV-specific human B cell clones expressing the intestinal homing receptor α4β7+ compared with that of randomly selected B cells. The relative frequency of three dominant VH gene segments in RV-specific intestinal homing B cells (▨) is plotted in comparison to the frequency of use of these VH gene segments in randomly selected (▪) or RV-specific (▦) systemic IgD B cells or randomly selected intestinal homing B cells (□). The frequency of use of the VH3–23 segment (the dominant segment in randomly selected systemic blood B cells) is shown for comparative purposes. Eighty-three clones were analyzed for randomly selected systemic cells, 18 for randomly selected IgD4β7+ intestinal homing cells, 18 for RV-specific systemic IgD cells, and 33 for RV-specific IgD4β7+ intestinal homing B cells. The 83 individual VH sequences from randomly selected systemic B cells and 18 individual VH sequences from RV-specific systemic IgD are published as supplemental data in Ref. 7 . Comparisons in proportions for this figure were performed by using the Fisher’s exact test statistic. When proportions of VH gene segment usages were compared between randomly selected peripheral B cells and randomly selected intestinal homing B cells using this method, the p value was 0.032. Comparison of proportions between randomly selected intestinal homing B cells and RV-specific intestinal homing B cells resulted in a p value of 0.002.

Close modal

The VH1–46 gene segment, rarely used in randomly selected B cells (0 of 83 randomly selected adult B cells in our study), was highly overrepresented in RV-specific B cells in either the systemic memory (6 of 18 clones) or intestinal homing (13 of 33 clones) subsets (Fig. 2). Although the VH gene segment usage of randomly selected intestinal homing B cells was biased toward the VH1 and VH4 gene segment families, VH1–46 was not used in any of the 18 randomly selected intestinal homing B cell clones that were analyzed. This difference in VH gene segment usage between randomly selected and RV-specific intestinal homing memory B cells was statistically significant (p = 0.002). Overall, these data confirm our previous finding that the VH1–46 gene segment is the dominant gene determinant of RV-specific Ab responses. Additional dominant VH gene segments that we observed previously in RV-specific B cells, such as VH4–31 and VH4–39 (7), were also identified in RV-specific intestinal homing B cells, suggesting a strong antigenic selection bias for these VH segments (Fig. 2).

D segments were assignable in 3 of 13 VH1–46 gene segments (23%) in intestinal homing RV-specific B cells compared with 3 of 6 VH1–46 segments (50%) in systemic RV-specific memory B cells (Table II). For non 1–46 VH gene segments, D segment assignment was possible in 67% of 12 sequences from systemic RV-specific memory B cells, 72% of 18 sequences from intestinal homing randomly selected memory B cells, and 47% of 20 sequences from intestinal homing RV-specific B cells. We found that more than one-half of all clones in RV-specific intestinal homing and RV-specific systemic memory B cells with assignable D segments belonged to the D3 family. This D3 bias was also observed in randomly selected intestinal homing B cells (25% of all assignable clones expressed this D segment). Therefore, we could not determine whether the D3 gene segment was selected because of Ag specificity. We observed a uniform bias toward the JH4 family in all populations, which corresponds to the bias seen in nonspecific adult B cells, as shown previously (7).

Table II.

Mutational analysis of H chain CDRs and FRs of B cells from healthy adults sorted by phenotype

Type of B CellPhenotypic MarkersDonors AnalyzedVH Gene Segment (n)aMean % Nucleotide Change from GermlineMean No. of Replacement (R) or Silent (S) Mutations in Indicated RegionsCDR3 Mutations
CDR1 and 2 (combined)FR1b, 2, and 3 (combined)D segment not assignable (%)JH segment mutated (%)CDR3 length (amino acids)
RSRS
Systemic RV-specific memory B cells RV+ IgD Non 1–46 (12) 8.6c 4.7 1.9 9.8 7.5 67 25 15 
   1–46 (6) 5.7d 1.8 1.8 4.2 50 83 13 
Intestinal homing randomly selected memory B cells IgD α4β7+ Non 1–46 (18) 5.3e 4.6 1.1 2.7 72 43 16 
Intestinal homing RV-specific memory B cells RV+ IgD α4β7+ Non 1–46 (20) 1.9ce 1.2 0.5 1.6 1.2 47 35 15 
   1–46 (13) 0.3d 0.3 0.2 23 62 14 
Type of B CellPhenotypic MarkersDonors AnalyzedVH Gene Segment (n)aMean % Nucleotide Change from GermlineMean No. of Replacement (R) or Silent (S) Mutations in Indicated RegionsCDR3 Mutations
CDR1 and 2 (combined)FR1b, 2, and 3 (combined)D segment not assignable (%)JH segment mutated (%)CDR3 length (amino acids)
RSRS
Systemic RV-specific memory B cells RV+ IgD Non 1–46 (12) 8.6c 4.7 1.9 9.8 7.5 67 25 15 
   1–46 (6) 5.7d 1.8 1.8 4.2 50 83 13 
Intestinal homing randomly selected memory B cells IgD α4β7+ Non 1–46 (18) 5.3e 4.6 1.1 2.7 72 43 16 
Intestinal homing RV-specific memory B cells RV+ IgD α4β7+ Non 1–46 (20) 1.9ce 1.2 0.5 1.6 1.2 47 35 15 
   1–46 (13) 0.3d 0.3 0.2 23 62 14 
a

The first 24 nt of FR1 were encoded by primers and therefore not analyzed for mutations.

b

VH1–46 was the dominant VH gene segment in RV-specific B cells. This segment did not occur in RV-unselected intestinal homing memory B cells.

c

p < 0.0001.

d

p = 0.002.

e

p = 0.003.

We also analyzed the Ab L chain variable gene sequences obtained from RV-specific α4β7+ B cells and compared their family distribution with that of systemic compartment RV-specific IgD B cells. We did not detect a difference in the VL family distribution between these populations. For VL sequence analysis, 46 total clones from 3 donors were analyzed for randomly selected systemic cells, 18 cells from 6 donors for randomly selected IgD4β7+ intestinal homing cells, 24 clones from 2 donors for RV-specific systemic IgD cells, and 21clones from 7 donors for RV-specific IgD4β7+ intestinal homing B cells. The 46 individual VL sequences from randomly selected systemic B cells and 24 individual VL sequences from RV-specific systemic IgD were published as supplemental data in Ref. 7 . The κλ L chain ratio of >2:1 and the exclusive use of the Vκ1, Vκ3, and Vκ4 families were observed in both populations. Vκ1 was overrepresented, with 40 and 47% in the RV-specific intestinal homing B cells and the systemic compartment RV-specific memory B cells, respectively.

We analyzed the frequency of somatic hypermutations in RV-specific memory B cells with or without expression of α4β7. We found, as expected, a high mutation frequency in systemic RV-specific IgD memory B cells. In these cells, the mean nucleotide change from germline sequences was 8.6% for Abs that did not use VH1–46 and 5.7% for those that did use the RV-immunodominant VH gene segment 1–46 (Table II). We performed an ANOVA to examine mean percent nucleotide change between RV-specific IgD, randomly selected IgD4β7+, and RV-specific IgD4β7+ B cells. The overall test for significance was statistically significant (p = 0.003). Table II shows the p values for comparisons of interest. Both unadjusted p values and p values adjusted for multiple comparisons (using the Tukey-Kramer adjustment) resulted in the same outcome. Both VH1–46- and non-VH1–46-containing Ab sequences from RV-specific IgD B cells that expressed the intestinal homing marker α4β7 demonstrated a significantly lower mutation frequency compared with RV-specific memory B cells from the systemic compartment (p ≤ 0.002).

We did not observe mutations in the CDR1 or CDR2 regions in the 13 sequences of RV-specific intestinal homing B cells that used the RV-immunodominant VH1–46 gene segment. The mean percent nucleotide change from germline sequences in the variable genes of these B cells was only 0.3%. Analysis of the frequency of mutations within the CDR3 regions of H chains is inherently difficult; however, we noted that D and JH segments were frequently mutated in RV-specific memory B cells in the systemic compartment and in RV-specific intestinal homing B cells (Table II). We found the lowest percentage of assignable D segments in intestinal homing RV-specific B cells with 23% (3 of 13 sequences) assignable in VH1–46 gene segments from those cells, suggesting an increase in CDR3 mutations in contrast to the unmutated CDRs 1 and 2. The CDR3 length did not differ between groups. These data imply that circulating human B cells that are particularly suitable for interaction with RV in the intestine (i.e., B cells that are both RV-specific and intestinal homing) do not possess somatic hypermutations in the CDR1 and CDR2 regions.

Alignment of the VH1–46 sequences revealed particular positions in the variable gene that were commonly mutated (Fig. 3,A). In 14 instances, the same position was mutated in more than one clone. In eight cases, the identical amino acid change was found in two clones and in two cases was found in three clones. The mutations tended to cluster in the CDR1 and CDR2 regions, but common framework mutations were also observed. These findings suggest that particular positions in the VH1–46-derived Abs contribute differentially to contact with Ag. We analyzed the mutations at the nucleotide level to determine whether or not the mutations were most common in previously defined mutational hot spots (WRGY and TA sequences). We found that 33% of observed mutations in RV-specific IgD4β7+ and 42% of observed mutations in RV-specific IgD 4β7-unselected B cells occurred in mutational hot spots, suggesting that the general process of somatic hypermutation in RV-specific B cells proceeds with an expected distribution of mutations in the Ab gene sequences (Fig. 3 D).

FIGURE 3.

Comparison of the RV-specific Ab sequences using VH1–46 isolated from B cells of the indicated subsets. A, The variable regions of the VH1–46 sequences obtained from systemic IgD-unselected, systemic IgD-negative, or the α4β7+ IgD-negative B cell populations are shown. Identical amino acids are shown in gray and mutations are shown in white. The stars indicate positions at which an amino acid was mutated in more than one single-cell clone; a star with a letter underneath indicates that the position was mutated to that amino acid in two clones. An underlined letter designates that the residue at that position was mutated to the indicated amino acid in three or more B cell clones. B, Alignment of the HCDR3 regions of the VH1–46 sequences obtained from the three B cell populations. Amino acids identical to the germline D or J segments of the individual clone are shown in gray, whereas the mutated amino acids are shown in white. C, Alignment of the framework regions (FR) 4 of the VH1–46 sequences obtained from the three B cell populations. Amino acids identical to the germline J segment of the individual clones are shown in gray and mutations are shown in white. D, Summary of total mutations and relation to mutational hot spots. The number and percentage of mutations in the indicated motif is shown for each of the cell populations.

FIGURE 3.

Comparison of the RV-specific Ab sequences using VH1–46 isolated from B cells of the indicated subsets. A, The variable regions of the VH1–46 sequences obtained from systemic IgD-unselected, systemic IgD-negative, or the α4β7+ IgD-negative B cell populations are shown. Identical amino acids are shown in gray and mutations are shown in white. The stars indicate positions at which an amino acid was mutated in more than one single-cell clone; a star with a letter underneath indicates that the position was mutated to that amino acid in two clones. An underlined letter designates that the residue at that position was mutated to the indicated amino acid in three or more B cell clones. B, Alignment of the HCDR3 regions of the VH1–46 sequences obtained from the three B cell populations. Amino acids identical to the germline D or J segments of the individual clone are shown in gray, whereas the mutated amino acids are shown in white. C, Alignment of the framework regions (FR) 4 of the VH1–46 sequences obtained from the three B cell populations. Amino acids identical to the germline J segment of the individual clones are shown in gray and mutations are shown in white. D, Summary of total mutations and relation to mutational hot spots. The number and percentage of mutations in the indicated motif is shown for each of the cell populations.

Close modal

The central finding of this study is that RV-specific memory B cells in the peripheral blood that exhibit an intestinal homing phenotype share the overrepresentation of the immunodominant VH gene segment VH1–46 with that in the systemic compartment. RV replication in vivo is restricted to enterocytes, and it is likely that most circulating RV-specific B cells originate from cells in the intestinal Peyer’s patches. Recent studies showed that RV can escape the gastrointestinal tract in children, resulting in antigenemia and possibly viremia (10). This occurrence would explain exposure of the systemic compartment to RV Ags. B cells migrating through peripheral blood back to the gut should express the intestinal homing receptor α4β7. Gonzales et al. (6) found a mean percentage of 1.9 of memory B cells that bound RV DLP by flow cytometry analysis of peripheral blood cells obtained from healthy adults in Columbia, South America. In that study, the mean frequency of IgD B cells that were both RV-specific and expressed α4β7 was 3.3% in “large-” and 0.6% in “small-” sized memory B cells in acutely infected adults. We did not differentially sort memory B cells based on size in this study, but the mean frequency of RV-specific B cells was lower at 0.04% in our study population of healthy blood donors from the United States. This difference might be accounted for by a less frequent exposure to RV in North American adults or by differences in technique. However, we confirmed the finding of Gonzales et al. (6) that a significant proportion of RV-specific memory B cells expressed the intestinal homing marker α4β7. We did not determine the isotype expressed on RV-binding B cells in this study. Expansion of B cells in culture using CD154-expressing feeder cells and cytokines such as IL-2, IL-4, and T cell-replacing factor as in our system can induce isotype switching. However, when we analyzed the isotype from supernatants of expanded B cell clones, IgD+ B cells yielded a majority of clones producing IgM while IgD B cells yielded clones that produced either IgG or IgA (data not shown). Previous studies showed that approximately two-thirds of surface IgA+ memory B cells are α4β7+, whereas approximately one-third of circulating IgG+ B cells express α4β7 (3).

Ab gene analysis at the single-cell level revealed that the Ab repertoire in randomly selected intestinal homing B cells was VH4 dominated. We found a bias toward the VH4 gene segment family in randomly selected intestinal homing B cells (44% of all clones). We observed a similar bias toward the D3 and JH4 gene segment families and a similar distribution of VL gene segments in all studied groups. In contrast, when we analyzed individual VH gene segment usage, we found a significant difference between randomly selected systemic and α4β7+ memory B cells. These data suggest that intestinal homing B cells express a biased Ab repertoire and confirm that we analyzed two distinct B cell populations following physical selection based on immunophenotyping. We are not aware of a previous analysis of VH family distribution in human α4β7+ memory B cells, but McCabe et al. (11) analyzed VH usage in human intestinal B cells from colonoscopic biopsies and also found a VH4 dominance. In contrast, the VH gene repertoire in human intestinal plasma cells is VH3 dominated (12, 13). Human Ab responses to capsular polysaccharides are encoded by the VH3 gene segment family (reviewed in Ref. 14) and random B cells in the systemic compartment are uniformly VH3-biased (15, 16, 17, 18).

RV-specific intestinal homing B cells demonstrated a VH1/VH4 gene family bias very similar to that of RV-specific systemic memory B cells. We confirmed the strong VH1/VH4 gene family bias in RV-specific B cells identified in our previous work (7). We also confirmed that VH1–46 is the dominant gene segment in human B cells responding to RV. We found that VH1–46 was significantly overrepresented in RV-specific memory B cells when compared with randomly selected memory B cells, whether of the intestinal homing phenotype or not. In our previous study of RV-specific B cells, we found that VH1–46 was the most common VH gene segment in RV-specific B cells in the systemic compartment and showed that a recombinant Ab specified by the VH1–46 gene segment Ab binds to RV VP6 (7, 8). It was of interest to find that the systemic and mucosal homing B cells shared a similar repertoire.

We used the term “systemic” for α4β7-unselected B cells. The only difference in sorting between the RV-specific intestinal homing B cells and RV-specific systemic memory B cells was that the latter cells were α4β7 unselected. We are confident that these are two different cell populations, since we found surprisingly a statistically significant difference in the frequency of somatic hypermutations. Some overlap between α4β7-unselected and α4β7+ B cells cannot be ruled out, however. As shown in Table I, we found that a mean of 10% of RV-specific B cells also exhibited the intestinal homing phenotype. Therefore, the vast majority of α4β7-unselected B cells did not exhibit an intestinal homing phenotype.

RV-specific intestinal homing B cells exhibited a low frequency of somatic hypermutation, particularly in cells using the RV-specific immunodominant gene segment VH1–46. We were surprised to find a statistically significant lower frequency of somatic hypermutations in RV-specific intestinal homing B cells using the VH1–46 gene segment (0.3% nucleotide change from germline) compared with RV-specific memory B cells in the systemic compartment (5.7% nucleotide change from germline). Previous studies demonstrated a mutation frequency in human intestinal plasma cells secreting IgA and IgM that exceeded the level of somatic hypermutation of VH genes carried by human memory B cells (12). In contrast, we sorted intestinal homing B cells before their differentiation into plasma cells. During the in vitro stimulation that we used, single B cells were expanded and differentiated into cells that secreted Ig to confirm the RV specificity of the physical sorting methods. It is not surprising that mutations were not induced by our in vitro stimulation procedure, since it is well established that somatic hypermutations are exceedingly difficult to induce ex vivo. It is unlikely that our method of culture only selected for the outgrowth of B cells lacking mutations since the same culture system generated B cell clones from systemic memory B cells that were highly mutated.

When we separately analyzed intestinal homing B cell clones that used the immunodominant gene segment VH1–46, we found that these clones in particular exhibited a low frequency of somatic hypermutation. The VH1–46 gene segments did not possess mutations in their CDR1 or 2 regions. This finding is remarkable given the strong association of this VH gene segment with RV specificity independent of D, JH, and JL segments in this study and our previous studies, and suggests that the CDR1 or CDR2 regions may play a dominant role in binding of these Abs. Therefore, the high percentage of RV-specific intestinal homing B cells that use VH1–46 (39%), and the lack of mutations in that gene segment, accounted for the overall low frequency of somatic hypermutation in the study. Systemic RV-specific memory B cells used the VH1–46 gene segment with a similar frequency (33%), but only 50% of these clones lacked mutations in their CDR1 and 2 regions. Also, the overall frequency of somatic hypermutations was higher in the systemic RV-specific memory B cell groups, both VH1–46 and non-VH1–46, than in the intestinal homing B cells. It should be noted that an element of defining the intestinal homing cells was an IgD phenotype, so that the intestinal homing cells may be considered a subset of memory cells as evidenced by class switch recombination. Therefore, the lack of mutations in these cells is striking and suggests that they may have expanded outside of a germinal center. A possible reason for the low mutational frequency in RV-specific B cell clones using VH1–46 could be that the germline sequence of this immunodominant gene segment specifies CDR1 and CDR2 regions providing an optimal interacting surface for the Ag due to excellent shape complementarity or other biochemical features and therefore does not require mutations to bind efficiently. The differential frequency of mutations in VH1–46 in the systemic and intestinal homing compartments, however, remains unexplained. The molecular mechanisms mediating somatic hypermutation, such as activation-induced cytidine deaminase and error-prone DNA polymerases, have only recently been discovered. It is possible that the expression of these molecules is differentially regulated in the intestinal milieu. Alternatively, it may be that α4β7+ B cells in the intestine do not enter germinal centers because of surface interactions mediated by the integrin during trafficking. We also speculate that the interaction of α4β7 on B cells with mucosal addressin cell adhesion molecule 1 might cause an enhanced avidity of RV-specific B cells in the intestine that otherwise have low binding strength for Ag due to low-affinity Fv-mediated interactions with RV-infected cells. Such avidity effects might alter the survival of B cells displaying Ag receptors that are of low affinity.

In summary, we found that the mean frequency of RV-specific B cells in healthy blood donors was 0.6%, and the frequency of RV-specific intestinal homing B cells ranged between 0.01 and 0.1%. Thus, intestinal homing B cells represent a minority of the RV-specific B cells circulating in the peripheral blood. Intestinal homing B cells demonstrated a stronger VH4 family bias than systemic compartment B cells and a frequent usage of gene segment VH4–39. RV-specific intestinal homing B cells used the immunodominant gene segment VH1–46 at a high frequency and were not mutated, suggesting an optimal fit for interaction with Ag. These data confirm that VH1–46 is an immunodominant gene segment in human effector B cells responding to RV and show that the systemic and intestinal homing B cells share features of a common RV-specific repertoire.

The authors have no financial conflict of interest.

We thank Catherine Allen and James Price of the Nashville Veterans Affairs Hospital for excellent flow cytometry technical support; DNAX for use of the CD154 expressing cell line; Rudolf H. Zubler for the EL4-B5 cell line; and the National Cancer Institute BRB Preclinical Repository (Rockville, MD) for recombinant human IL-2. Eugene Butcher graciously provided the α4β7-specific Ab. We thank Elizabeth Bures, Nicholas Shinners, and Frances House for excellent technical support in B cell isolation and cloning and Ninguo Feng and Dino Feigelstock for VLP preparation.

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 grants from the National Institute of Child Health and Human Development and RO1 AI21362 (R01 HD36311), National Institute of Allergy and Infectious Diseases (R01 AI-57933), the Vanderbilt National Institutes of Health General Clinical Research Center Grant M01 RR00095. J.-H.W. was a 1999–2001 Fellow of the Deutsche Forschungsgemeinschaft (WE 2405/1-1) and the 2001–2002 Bayer Harold Neu Postdoctoral Fellow of the Infectious Diseases Society of America. N.L.K. and A.L.B. were supported by the National Institutes of Health Training Grant T32 AI-07611. DNA sequence analysis was performed at the Vanderbilt-Ingram Cancer Center sequencing core facility, funded by Cancer Center Support Grant IP30 CA68485. Support was provided by a grant from the National Institute of Diabetes and Digestive and Kidney Diseases (P30 DK56339).

3

Abbreviations used in this paper: RV, rotavirus; VLP, virus-like particle; DLP, double-layered particle; FR, framework region.

1
Kuklin, N. A., L. Rott, N. Feng, M. E. Conner, N. Wagner, W. Müller, H. B. Greenberg.
2001
. Protective intestinal anti-rotavirus B cell immunity is dependent on α4β7 integrin expression but does not require IgA antibody production.
J. Immunol.
166
:
1894
.
2
Dunn-Walters, D. K., P. G. Isaacson, J. Spencer.
1997
. Sequence analysis of human IgVH genes indicates that ileal lamina propria plasma cells are derived from Peyer’s patches.
Eur. J. Immunol.
27
:
463
.
3
Brandtzaeg, P., I. N. Farstad, G. Haraldsen.
1999
. Regional specialization in the mucosal immune system: primed cells do not always home along the same track.
Immunol. Today
20
:
267
.
4
Rott, L. S., M. J. Briskin, E. C. Butcher.
2000
. Expression of α4β7 and E-selectin ligand by circulating memory B cells: implications for targeted trafficking to mucosal and systemic sites.
J. Leukocyte Biol.
68
:
807
.
5
Williams, M. B., J. R. Rosé, 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
.
6
Gonzales, A. M., M. C. Jaimes, I. Cajiao, O. L. Rojas, J. Cohen, P. Pothier, E. Kohli, E. C. Butcher, H. B. Greenberg, J. Angel, M. A. Franco.
2003
. Rotavirus-specific B cells induced by recent infection in adults and children predominantly express the intestinal homing receptor α4β7.
Virology
305
:
93
.
7
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 immunodominant variable gene repertoires.
J. Immunol.
171
:
4680
.
8
Weitkamp, J.-H., N. Kallewaard, K. Kusuhara, D. Feigelstock, N. Feng, H. B. Greenberg, J. E. Crowe, Jr.
2002
. 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
.
9
Weitkamp, J.-H., J. E. Crowe, Jr.
2001
. Blood donor leukocyte reduction filters as a source of human B lymphocytes.
BioTechniques
31
:
464
.
10
Blutt, S. E., C. D. Kirkwood, V. Parreno, K L. Warfield, M. Ciarlet, M. K. Estes, K. Bok, R. F. Bishop, M. E. Conner.
2003
. Rotavirus antigenaemia and viraemia: a common event?.
Lancet
362
:
1445
.
11
McCabe, R. P., W. L. Carroll, M. Egan, S. M. Cohn, M. Peters.
1994
. Immunoglobulin variable region usage in human intestinal B lymphocytes.
Clin. Immunol. Immunopathol.
71
:
240
.
12
Fischer, M., R. Küppers.
1998
. Human IgA- and IgM-secreting intestinal plasma cells carry heavily mutated VH region genes.
Eur. J. Immunol.
28
:
2971
.
13
Scamurra, R. W., D. B. Nelson, X. M. Lin, D. J. Miller, G. J. Silverman, T Kappel, J. R. Thurn, E. Lorenz, A. Kulkarni-Narla, E. N. Janoff.
2002
. Mucosal plasma cell repertoire during HIV-1 infection.
J. Immunol.
169
:
4008
.
14
Pirofski, L..
2001
. Polysaccharides, mimotopes and vaccines and encapsulated pathogens.
Trends Microbiol.
9
:
445
.
15
Brezinschek, H. P., R. I. Brezinschek, P. E. Lipsky.
1995
. Analysis of the heavy chain repertoire of human peripheral B cells using single-cell polymerase chain reaction.
J. Immunol.
155
:
190
.
16
Suzuki, I., L. Pfister, A. Glas, C. Nottenburg, E. C. B. Milner.
1995
. Representation of rearranged VH gene segments in the human adult antibody repertoire.
J. Immunol.
154
:
3902
.
17
Kraj, P., S. P. Rao, A. M. Glas, R. R. Hardy, E. C. B. Milner, L. E. Silberstein.
1997
. The human heavy chain Ig V region gene repertoire is biased at all stages of B cell ontogeny, including early pre-B cells.
J. Immunol.
158
:
5824
.
18
Rao, S. P., J. M. Riggs, D. F. Friedman, M. S. Scully, T. W. LeBien, L. E. Silberstein.
1999
. Biased VH gene usage in early lineage human B cells: evidence for preferential Ig gene rearrangement in the absence of selection.
J. Immunol.
163
:
2732
.