Antibodies Encoded by FCRL4-Bearing Memory B Cells Preferentially Recognize Commensal Microbial Antigens

Yanling Liu, Jonathan R. McDaniel, Srijit Khan, Paolo Campisi, Evan J. Propst, Theresa Holler, Eyal Grunebaum, George Georgiou, Gregory C. Ippolito and Götz R. A. Ehrhardt
J Immunol June 15, 2018, 200 (12) 3962-3969; DOI: https://doi.org/10.4049/jimmunol.1701549

Abstract

FCRL4, a low-affinity IgA Ab receptor with strong immunoregulatory potential, is an identifying feature of a tissue-based population of memory B cells (Bmem). We used two independent approaches to perform a comparative analysis of the Ag receptor repertoires of FCRL4+ and FCRL4− Bmem in human tonsils. We determined that FCRL4+ Bmem displayed lower levels of somatic mutations in their Ag receptors compared with FCRL4− Bmem but had similar frequencies of variable gene family usage. Importantly, Abs with reactivity to commensal microbiota were enriched in FCRL4+ cells, a phenotype not due to polyreactive binding characteristics. Our study links expression of the immunoregulatory FCRL4 molecule with increased recognition of commensal microbial Ags.

Introduction

Memory B cells (Bmem) are pivotal to the establishment and maintenance of humoral protection from re-encountered pathogens, both in the context of natural infection and vaccination. Following exposure to Ag, Bmem proliferate and differentiate into Ab-secreting plasma cells more rapidly than naive B cells (1). The importance of pathogen-specific Abs is emphasized by observations illustrating that neutralizing Ab responses is an immune correlate of many successful vaccines (2–4). Furthermore, persisting Bmem also provide a diversified repertoire of Abs able to target escape mutants of pathogens that are no longer effectively engaged by the initial Ab response (5).

Various studies have demonstrated the heterogeneous nature of the Bmem compartment in mice and humans. In mice, functionally distinct subpopulations of Bmem have been defined by differential expression of Ab isotype (6–8) and by differential expression of the CD73, CD80, and PD-L2 Ags (9, 10). In humans, circulating Bmem are typically characterized by expression of CD27 Bmem (11, 12); several reports demonstrated subpopulations of Bmem in blood and tissue that lacked expression of the CD27 Ag (13–15).

The Fc receptor–like (FCRL) family of immunoregulators are preferentially expressed on B lineage cells and display a Bmem-centric pattern of expression (16). FCRL4 characterizes a morphologically and functionally distinct population of tissue-based Bmem with a distinctive gene expression profile (14, 17, 18). The extracellular domain of FCRL4 functions as a low-affinity receptor for IgA (19), and its intracellular domain exhibits potent regulatory activity on Ag receptor signaling (20–22). This inhibitory activity is mediated by recruitment of the SHP1 and SHP2 tyrosine phosphatases to ITIM consensus sequences (20, 22). Importantly, dysregulation of FCRL4-bearing Bmem was observed in the context of HIV immunopathology, in which Abs targeting the HIV gp120 envelope protein were enriched in the FCRL4+ population (23–27). Involvement of FCRL4+ Bmem was also reported in the immunopathology of malaria (23) and rheumatoid arthritis (28). These unique features prompted us to investigate the Ab repertoire of FCRL4+ Bmem in healthy individuals. We observed that Abs from FCRL4+ Bmem had lower levels of somatic mutations than Abs from FCRL4− Bmem while displaying comparable variable gene usage. Importantly, Abs derived from FCRL4+ Bmem showed increased reactivity to microbiota, a characteristic that was not accompanied by autoreactive or polyreactive binding characteristics relative to Abs from FCRL4− Bmem. Our study links the cell surface expression of the immunoregulatory FCRL4 molecule to increased reactivity to commensal Ags.

Materials and Methods

Abs and reagents

Abs to CD19 (clone HIB-19), CD38 (clone HIT-2), IgD (clone 1A6-2) and IgM (clone G20-127), CD3 (clone SK7), and Igκ (clone G20-193) were obtained from BD Biosciences (San Jose, CA). Abs to FCRL4 (clone 1A3) were provided by Genentech (South San Francisco, CA). Protein A-Sepharose was obtained from Amersham Biosciences (Piscataway, NJ). Polyclonal HRP-coupled rabbit anti-human Ig Abs were obtained from Jackson ImmunoResearch (West Grove, PA).

Cell lines and primary cells

HEK293T cells and HEp-2 cells were grown in DMEM, supplemented with 10% FBS and 100 U/ml penicillin/streptomycin. Cells were grown in humidified atmosphere at 37°C and 5% CO2. Tonsillar tissue from pediatric patients undergoing routine tonsillectomy was obtained from The Hospital for Sick Children (Toronto, ON, Canada) with informed consent, according to the Declaration of Helsinki.

Generation of mAbs

Cell suspensions of tonsil tissue were generated by tissue mincing using 70 μm steel mesh. Mononuclear cells were prepared by density gradient centrifugation with lymphocyte separation medium. Individual class–switched FCRL4+ or FCRL4− Bmem (CD19+CD38−IgD−IgM−) were FACS sorted into 96-well PCR plates containing 10 μl of RT-PCR catch buffer supplemented with RNasin (Promega, Madison, WI) (for gating strategy, see Supplemental Fig. 1). FCRL4 staining was performed using anti-FCRL4 clone 1A3, owing to the stronger signal compared with anti-FCRL4 clone 4-2A6, as previously described (14). The plates were immediately sealed and frozen at −80°C. Single-cell RT-PCR was performed as described (29). Primary PCR reactions were performed with OneStep RT-PCR (Qiagen, Hilden, Germany) and secondary PCR reactions with KOD enzymes (EMD Millipore, Billerica, MA). Amplified H and L chain sequences were verified by DNA sequencing, and sequences were annotated using the international ImMunoGeneTics platform (30). Recombinant mAbs were generated by transient transfection of H and L chain–containing vectors into HEK293T cells using the polyethylenimine method (31). Secreted Abs were purified from culture supernatant using Protein A-Sepharose and dialyzed against PBS.

VH repertoire sequencing

FCRL4+ or FCRL4− Bmem were pelleted by centrifugation (300 × g for 10 min at 4°C) and resuspended in 1 ml TRIzol reagent (Thermo Fisher, Waltham, MA). RNA was extracted from the solution using an RNeasy system (Qiagen), according to the manufacturer’s instructions. First strand cDNA was generated from 500 ng of total RNA with SuperScript II (Thermo Fisher). VH repertoires were amplified via PCR using gene-specific primers (32) and were sequenced on 2 × 300 Illumina MiSeq. Ab lineages were defined by grouping sequences on ≥90% nt identity of the CDRH3. To measure the differences between FCRL4+ and FCRL4− cells, properties were averaged by equally weighting all lineages, as opposed to averaging all sequences, which would intrinsically weight the properties toward lineages with higher read counts.

VH:VL repertoire sequencing

FCRL4+ or FCRL4− Bmem were resuspended in PBS and coemulsified with oligo d(T)25 magnetic beads (NEB) and lysis buffer using a custom-designed axisymmetric flow–focusing device as described previously (33, 34). Briefly, magnetic beads were used to capture mRNA from single B cells, after which the beads were extracted, washed, resuspended in an RT-PCR mix containing VH and VL linkage primers (33), emulsified, and subjected to RT-PCR using the following conditions: 30 min at 55°C followed by 2 min at 94°C; 4 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 2 min; 4 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 2 min; 32 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 2 min; 72°C for 7 min; held at 4°C. The VH:VL–stitched amplicons were then extracted from the emulsions, amplified via seminested PCR, and gel-purified. Finally, the amplicons were prepared for Illumina MiSeq by appending Illumina compatible barcodes. The raw next generation sequencing data reported in this paper will be made available through the Sequence Read Archive.

Bioinformatic analysis

Raw 2 × 300 Illumina reads were stitched using paired-end read merger software (VH repertoire) (35) or trimmed to remove low-quality sequences using Trimmomatic (VH:VL repertoire) (36), filtered for sequence quality, and submitted to MiXCR for CDR3 identification and gene annotation (37). Sequences with ≥2 reads were grouped based on 90% CDR-H3 nucleotide identity using USEARCH (version 7.0). The dominant sequence in each group by read count was used to calculate CDR length, hydrophobicity (Kyte-Doolittle index) (38), V gene somatic hypermutation rate, and gene usage frequency.

Ab-binding assays

Commensal microbiota communities (MET1) were heat inactivated (56°C, 30 min) prior to resuspension in H2O and plated in a volume of 100 μl/well at an OD of 0.2. DNA was coated at a concentration of 10 μg/ml using the DNA coating solution (Thermo Fisher Scientific) overnight at room temperature, and insulin and LPS were coated at 5 μg/ml overnight at 4°C. Plates were incubated with 1 μg/ml primary Abs overnight at 4°C. Bound Abs were detected with HRP-labeled rabbit anti-human IgG and TMB ELISA substrate (Thermo Fisher Scientific). Serial dilutions of normal human serum were used as standard samples, and PBS was used as negative control. OD readings were normalized to standards of normal human serum to account for interplate variation. Each Ab-binding value represents the mean of at least three independently performed experiments. Moderate binding was defined as OD > mean plus 5 × SD of negative control and strong binding as OD > 4 × mean of negative control.

Statistical analysis

Statistical significance for VH and VL mutation analyses was determined using Mann–Whitney U tests. Statistical significance for VH and VL gene usage and Ab-binding frequency to commensal microbiota, HEp-2, LPS, insulin, and DNA was determined using Fisher exact tests. Correlations were determined using Spearman rank-order correlation test. Statistical analyses were performed using Prism 5 software. The p values <0.05 were considered statistically significant.

Supplementary material

Supplemental Fig. 1 shows the gating strategy used to isolate FCRL4+ and FCRL4− Bmem. Supplemental Fig. 2 shows L chain gene analyses of paired VH:VL Ag receptor sequencing data as well as J gene usage analyses of all amplified Ag receptor H and L chain sequences. Supplemental Fig. 3 depicts strongly reactive, moderately reactive, and nonreactive mAbs from FCRL4+ and FCRL4− Bmem to commensal microbiota and HEp-2 lysates. Supplemental Fig. 4 shows individual Ag-binding data of all amplified Abs.

Results

Reduced frequency of somatic mutations in FCRL4+ Bmem

Expression of FCRL4 defines a morphologically and functionally distinct population of human Bmem. However, little is known about the Abs encoded by these cells in healthy individuals. To gain further insight into the characteristics of these Abs, we isolated FCRL4+ and FCRL4− Bmem from four independent tonsil samples and performed high-throughput sequencing on the H chain Ag receptor transcripts of 109,000 FCRL4+ Bmem yielding 31,905 lineages and 190,000 FCRL4− Bmem yielding 31,618 lineages. To determine whether the FCRL4 phenotype influenced V gene usage, we calculated the correlation between FCRL4+ and FCRL4− Bmem V gene frequency (ρ = 0.96; Fig. 1A, filled black symbols) and between the V gene family frequency (ρ = 0.96, Fig. 1A, open red symbols). The high degree of correlation between FCRL4+ and FCRL4− Bmem datasets suggests that the generation of the FCRL4+ phenotype is independent of the V gene identity. However, the FCRL4+ Bmem population exhibited a small but statistically significant increase in CDR-H3 length (Fig. 1B) and a greatly reduced number of somatic mutations incorporated into the V region (Fig. 1C, left panel), whereas CDR-H3 hydrophobicity was similar between FCRL4+ and FCRL4− Bmem (Fig. 1C, right panel). The small increase in CDR3 length, similar hydrophobicity, and comparable variable gene usage was also observed for L chain sequences (Supplemental Fig. 2). Similarly, J gene sequences were also comparable between FCRL4+ and FCRL4− Bmem (Supplemental Fig. 2). Additionally, we performed a comparative repertoire analysis of paired VH:VL sequences using 1628 FCRL4+-paired lineages and 1022 FCRL4−-paired lineages. This analysis showed that the FCRL4+ Bmem population displayed a reduced number of somatic mutations in both the H and L chain sequences compared with the FCRL4− Bmem population (Fig. 1D). These experiments show that expression of the FCRL4 immunoregulator identifies a subpopulation of human Bmem with reduced levels of somatically mutated Ag receptor genes.

FIGURE 1.
FIGURE 1.

Repertoire analysis of FCRL4+ and FCRL4− Bmem. FCRL4+ and FCRL4− Bmem were grouped into lineages according to the nucleotide identity of the CDR-H3. Only the dominant sequence by read count was used to represent each lineage in the following panels. (A) FCRL4+ and FCRL4− Bmem populations display similar VH gene usage frequencies. Spearman rank-order correlation coefficient is indicated. Filled black symbols indicate correlation between FCRL4+ and FCRL4− Bmem V gene frequency. Open red symbols indicate correlation between the V gene family frequency. (B) Increased CDR-H3 length in FCRL4+ Bmem. Numbers indicate amino acid (aa) length, and calculated means are indicated. (C) Decreased frequency of somatic mutations in CDR-H3 of FCRL4+ Bmem (median 2.1% versus 3.4%) (left panel). Hydrophobicity was calculated using the Kyte-Doolittle index and was found to be comparable between both populations (right panel). (D) Decreased frequency of somatic mutations in CDR3 regions of paired H and L chain sequences of FCRL4+ Bmem. Statistical significance was determined using the Mann–Whitney U test (B–D) and is indicated by an asterisk: *p < 0.05.

Reduced somatic mutations of VH4-34 sequences in FCRL4+ Bmem

A recent study investigating the Ab repertoires of individuals deficient in MYD88 and IRAK4 revealed that unmutated VH4-34 H chain sequences were enriched for reactivity to commensal microbiota. Specifically, it was reported that the VH4-34 sequence motif AVY located in the FR1 region and the N-linked glycosylation motif NHS located in the CDR-H2 contribute to commensal reactivity. These two sites were frequently found unmutated in this patient cohort (39). Although we observed VH4-34 IgG Bmem at similar frequencies (611 FCRL4+ and 590 FCRL4− lineages both representing 1.9% of their respective repertoires), the frequency of mutations leading to amino acid substitutions in the AVY–NHS motifs was higher in FCRL4− cells compared with FCRL4+ Bmem (Fig. 2). These findings suggest that Abs originating from FCRL4+ Bmem may exhibit reactivity to the commensal microbiome.

FIGURE 2.
FIGURE 2.

Increased frequency of unmutated VH4-34 sequences in class-switched IgG FCRL4+ Bmem. VH4-34 sequences were analyzed for mutations present in the AVY–NHS motifs. Numbers of mutated or unmutated VH4-34 genes are indicated. Statistical significance was determined using Fisher exact test.

Abs encoded by FCRL4+ Bmem preferentially bind to commensal Ags

In a separate series of experiments, we generated a set of 212 mAbs by single-cell RT-PCR (FCRL4+: 101 Abs and FCRL4−: 111 Abs) from three additional, independent tonsil samples. The most abundant sequences used VH3 and the VK1 family genes (Fig. 3A). The preferential use of these variable gene families is consistent with studies on mucosal plasmablasts (40). VH and VL gene usage was comparable between the two populations, reflecting the distribution we observed in our preceding repertoire-sequencing experiments. Analysis of V gene somatic hypermutation frequencies indicated significantly fewer somatic mutations in both H and L chain sequences of Ag receptors cloned from FCRL4+ Bmem (Fig. 3B), consistent with the V gene repertoire data above. As expected, the frequency of replacement to silent (R/S) mutations was increased in the complementarity-determining regions compared with framework regions (Fig. 3C). R/S ratios appeared increased in FCRL4+ Bmem, although, with the exception of sequences encoding L chain CDR2 regions, this did not reach statistical significance.

FIGURE 3.
FIGURE 3.

Reduced levels of somatic mutations in FCRL4+ Bmem. H chain (VH) and L chain (VL) sequences amplified from individual Bmem were analyzed for (A) variable gene usage and (B) presence of somatic mutations. Sequences obtained from FCRL4+ Bmem are indicated by closed circles and sequences obtained from FCRL4− Bmem by open circles. Mean values are indicated by horizontal bars. (C) Analysis of replacement and silent mutation frequencies. Calculated R/S values are indicated. Statistical significance was determined using Fisher exact test (A and C) and Mann–Whitney U test (B). Statistical significance is indicated by asterisks: *p < 0.05, **p < 0.001.

Increased levels of unmutated AVY–NHS motifs within the VH4-34 variable genes from FCRL4+ Bmem suggested the possibility for commensal-reactive, autoreactive, or polyreactive properties. This prompted us to investigate the underlying Ag-binding characteristics of mAbs from FCRL4+ and FCRL4− Bmem with respect to Ags reflecting these binding characteristics. Reactivity to microbial Ags was determined by recognition of heat-inactivated MET-1 cultures, a community of 33 isolates of human commensal bacteria used to simulate the commensal microbiome (41). Although MET-1 provides a complex mixture of epitopes, it represents only a fraction of the total complexity provided by the entire human microbiome, thus potentially resulting in moderate Ag recognition of reactive Ab clones. Self-reactivity was assessed by recognition of HEp-2 cell lysates (42, 43). Polyreactivity of Abs was tested using recognition of the LPS, insulin, and DNA Ags (43, 44). To assess Ab-binding over a broad spectrum of affinities, we defined Abs as moderately reactive if they exhibited binding of at least five standard deviations over the mean of the negative control and strongly reactive if they displayed binding of at least 4-fold the mean of the negative control (Supplemental Fig. 3). For all tonsil samples, we observed an increased frequency of commensal-reactive Abs derived from FCRL4+ cells (Fig. 4A, top panel). In contrast, no significant differences in Ag recognition were observed for binding to HEp-2 cell lysates (Fig. 4A). Similarly, we observed no differences in recognition of LPS, insulin, or DNA individually, except for increased DNA recognition of Abs from FCRL4− Bmem of tonsil 2 (Fig. 4A, bottom panel).

FIGURE 4.
FIGURE 4.

Increased binding of Abs from FCRL4+ Bmem to commensal Ags. (A) Abs amplified from Bmem of three independent tonsil samples were analyzed for reactivity to commensal microbiota, HEp-2 cell lysate, LPS, insulin, and DNA. The number of investigated Abs for each group is indicated in the center of each respective graph. Strongly reactive Abs are shown in red, moderately reactive Abs in gray, and nonreactive Abs in white. Statistical significance was determined using Fisher exact test and is indicated as p < 0.05 and n.s. for p > 0.05. (B) Correlation of binding to commensal microbiota with Ags indicating auto- or polyreactive binding. Values indicate relative OD450 following normalization to human serum standard, and correlation of binding was determined using Spearman rank-order correlation test. (C) Analysis of polyreactivity of Abs derived from FCRL4+ and FCRL4− Bmem. Polyreactivity was determined based on binding to two out of three (gray) or three out of three (black) Ags LPS, insulin, or DNA. Statistical significance was determined using Fisher exact test.

In subsequent analyses, we investigated whether the increased frequency of commensal-reactive Abs from FCRL4+ Bmem might reflect an increase in polyreactive clones. However, many of the commensal-reactive mAbs from FCRL4+ and FCRL4− Bmem did not bind to LPS, insulin, or DNA, and overall recognition of commensal microbiota correlated poorly with binding to insulin, LPS, or DNA (Fig. 4B). Similarly, binding of microbiome-reactive Abs from FCRL4+ and FCRL4+ Bmem showed weak correlation with HEp-2 lysate recognition (Fig. 4B). This contrasted with the strong correlation we observed between LPS and insulin recognition (Fig. 4B, bottom panel). Furthermore, we did not observe increased polyreactivity of commensal-reactive or HEp2-reactive Abs derived from FCRL4+ Bmem (Fig. 4C). To the contrary, overall polyreactivity defined as binding to at least two of the three tested Ags LPS, insulin, and DNA was increased in Abs derived from FCRL4− Bmem (Fig. 4C, Supplemental Fig. 4). These experiments indicate that cell surface expression of FCRL4 is a predictor of commensal microbiome-reactive cells.

FCRL4 expression, but not VH gene usage or somatic mutation frequency, indicates commensal reactivity

The increased frequency of commensal microbiome–reactive Abs among FCRL4+ Bmem prompted us to investigate whether the observed microbiome recognition was associated with increased usage of selective variable genes. Analysis of H chain variable gene usage of commensal-reactive clones demonstrated that the majority of sequences belonged to VH3 family members. However, we did not observe selective enrichment of a specific VH or VL gene family sequence among preferentially commensal-reactive Abs from FCRL4+ Bmem (Fig. 5A). Rather, the VH and VL gene usage of commensal-reactive clones recapitulated the distribution of the FCRL4+ Bmem repertoire (Fig. 3A).

FIGURE 5.
FIGURE 5.

FCRL4 expression, but not VH gene usage or somatic mutation frequency, indicates commensal reactivity. (A) Variable gene usage of combined moderate and strong commensal-reactive Abs for H chain (left) and L chain sequences (right) (FCRL4+, n = 83; FCRL4−, n = 49). Statistical significance was determined using Fisher exact test. (B) Analysis of somatic mutation frequencies for VH (left) and VL (right) sequences for commensal-reactive (red) and nonreactive Abs (black). Sequences from FCRL4+ Bmem are indicated by open circles and sequences from FCRL4− Bmem by open triangles. Mean values are indicated by horizontal bars, and numbers of commensal-reactive and nonreactive Abs for the various variable gene families are shown. Statistical significance for mutation frequencies was determined using a Mann–Whitney U test and is indicated by asterisks, *p < 0.05, **p < 0.01. Statistical significance for commensal-reactive frequencies was determined using Fisher exact test.

Next, we compared the somatic hypermutation frequency in the VH and VL gene families of commensal-reactive and nonreactive Abs. Increased commensal reactivity correlated with a decreased mutation frequency for all analyzed VH families (as well as most VL families) (Fig. 5B). However, commensal-reactive and nonreactive Abs did not appear to cluster by somatic hypermutation rate, indicating that the presence or absence of FCRL4 was a stronger indicator of commensal reactivity (Fig. 5B).

Discussion

FCRL4 is a low-affinity IgA receptor with strong immunoregulatory properties. In this study, we demonstrate that Abs from FCRL4+ Bmem have reduced levels of somatic mutations, and that the expression of FCRL4 in healthy individuals correlates with reactivity of the respective Abs to commensal microbiota.

Reduced levels of somatic mutations in FCRL4+ Bmem are in agreement with a report characterizing gp120-specific mAbs isolated from circulating, dysregulated tissue-like Bmem of HIV-viremic individuals, which are cells that share similarities with tonsillar FCRL4+ Bmem (45). Atypical circulating Bmem in individuals from malaria endemic regions also show similarities to FCRL4+ Bmem from tonsillar tissue (23); these cells display comparable variable gene usage relative to their classical Bmem counterparts but only trend toward reduced levels of somatic mutations (46). This contrasts with a recent report on the Ab repertoire of FCRL4+ Bmem in synovial fluid and tissue of rheumatoid arthritis patients in which variable gene usage analysis indicated overrepresentation of VH1-69 genes accompanied by underrepresentation of VH3-23 genes without differences in somatic mutation frequencies (28). The observed differences of the Ab repertoires could indicate potentially distinct contributions of FCRL4-bearing Bmem in healthy individuals and in the immunopathology of the various disorders. A recent report demonstrated that CD21low B lineage cells in blood represent recent emigrants from germinal centers, refractory to germinal center reentry (47). Transcriptome analysis of these cells revealed many shared characteristics with FCRL4+ Bmem of tonsillar origin, in addition to reduced levels of CD21 expression, including expression of CD11c, the Src family kinases Fgr and Hck and the SOX5 transcription factor. The reduced levels of somatic mutations we observed for FCRL4+ Bmem would be consistent with an inability of FCRL4-bearing cells to reenter germinal centers. However, unlike tonsillar FCRL4+ Bmem, circulating CD21low cells do not express FCRL4. Additional differences are found in increased levels of expression of transcription factors promoting plasma cell differentiation of circulating CD21low B lineage cells and in the absence of significantly reduced levels of somatic mutations in their Ag receptors compared with classical Bmem. It remains to be investigated whether tonsillar FCRL4+ Bmem will resemble CD21low B cells in blood more closely following exit from mucosal lymphoid tissue and entry into circulation.

For our experimental approach to determine microbiota recognition, we used a diverse community of microbiota representing human intestinal commensals. Our Ag binding studies indicate that FCRL4+ Bmem preferentially target microbial Ags. This Ag recognition pattern is unlikely the result of underlying polyreactive characteristics because it did not correlate with increased recognition of LPS, insulin, or DNA. A recent study of murine IgA–expressing B lineage cells revealed that, similar to this report, commensal reactivity was independent of the somatic mutation status of the cells (48). However, in contrast to our observations, the majority of commensal-reactive cells were polyreactive. This could reflect species differences or the more narrowly defined cell populations in this study, FCRL4+ versus FCRL4− Bmem. Alternatively, it could also reflect the more stringent definition of polyreactivity in our study, in which we defined polyreactivity as recognition of at least two out of the three Ags (LPS, insulin, and DNA), whereas the report by Bunker et al. (48) defined polyreactivity as recognition of at least two out of seven Ags (LPS, insulin, DNA, flagellin, albumin, cardiolipin, and keyhole limpet hemocyanin).

Tissue-like Bmem found in the blood of HIV-infected individuals share many characteristics with FCRL4+ Bmem of tonsillar tissue (24). However, B cell dysfunction is a well-recognized feature of the immunopathology of HIV infection (49). In contrast, tonsillar Bmem used in this study are more likely to represent healthy lymphocytes despite the caveat of recurrent tissue inflammation. It will be important to evaluate the binding characteristics of Abs isolated from FCRL4+ and FCRL4− Bmem of healthy individuals with specificities to known Ags (e.g., such as those elicited by common vaccinations). It will be equally important to determine potential functional differences in the plasma cell progeny of FCRL4+ and FCRL4− Bmem.

Tonsillar tissue provides a microenvironment that is rich in microbial Ags. The observed commensal reactivity of Abs derived from FCRL4-bearing Bmem is therefore likely to result in continuous Ag receptor signaling. Such continuous BCR signaling may explain the significant proportion of FCRL4+ Bmem in the G1 cell cycle phase in comparison with FCRL4− Bmem, which are primarily found in the G0 phase (17). The requirement for Ag in affinity maturation is well understood (50). In contrast, the contributions of direct Ag receptor signaling during cell fate determination in germinal centers are less clear (51, 52). Similarly, very limited information exists on factors regulating the expression of FCRL4 (17, 26, 53) and the generation of FCRL4-bearing Bmem, and no connection between FCRL4 expression and Ag receptor engagement has been established. Reduced levels of somatic mutations, the underlying mechanisms of which remain to be elucidated, characterize the Ab repertoire of FCRL4+ Bmem. It will be important to determine whether persistent Ag receptor signaling during affinity maturation is a regulatory factor in the generation of FCRL4+ Bmem and whether it contributes to the curbed replication history observed for dysregulated tissue-like Bmem in the immunopathology of HIV-1 (24) as well as to the reduced levels of somatic mutations that we observed in tonsillar tissues from healthy individuals.

FCRL4 expression defines a subpopulation of Bmem that is uniquely found in mucosa-associated lymphoid tissue in healthy individuals, but the specific function(s) of FCRL4 on these remains to be elucidated. Our study suggests a negative feedback scenario whereby FCRL4+ Bmem are subjected to constitutive Ag receptor signaling, which in turn will be inhibited by the interaction of FCRL4 and mucosal IgA immune complexes. FCRL4 may thus contribute to the maintenance of mucosal tolerance.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We are grateful to Dionne White and Joanna Warzyszynska for single-cell purification and to Dr. Emma Allen-Vercoe for generously providing commensal microbiota preparations. We thank the Genome Sequencing and Analysis Facility at the University of Texas at Austin for performing Illumina sequencing and Dr. Max D. Cooper (Emory University, Atlanta, GA) for critical reading of the manuscript.

Footnotes

  • This work was supported by Defense Threat Reduction Agency Contract HDTRA1-12-C-0105 (to G.G.), National Institutes of Health Grant 5R21AI119368 (to G.C.I.), and Canadian Institutes of Health Research Grants MOP-12614 and THA-11900 (to G.R.A.E.).

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    Bmem
    memory B cell
    FCRL
    Fc receptor–like
    R/S
    replacement to silent.

  • Received November 8, 2017.
  • Accepted April 9, 2018.

References

  1. ↵
    1. Tangye, S. G.,
    2. D. T. Avery,
    3. E. K. Deenick,
    4. P. D. Hodgkin
    . 2003. Intrinsic differences in the proliferation of naive and memory human B cells as a mechanism for enhanced secondary immune responses. J. Immunol. 170: 686–694.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Amanna, I. J.,
    2. M. K. Slifka
    . 2011. Contributions of humoral and cellular immunity to vaccine-induced protection in humans. Virology 411: 206–215.
    OpenUrlCrossRefPubMed
    1. Ahmed, R.,
    2. D. Gray
    . 1996. Immunological memory and protective immunity: understanding their relation. Science 272: 54–60.
    OpenUrlAbstract
  4. ↵
    1. Tangye, S. G.,
    2. D. M. Tarlinton
    . 2009. Memory B cells: effectors of long-lived immune responses. Eur. J. Immunol. 39: 2065–2075.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Purtha, W. E.,
    2. T. F. Tedder,
    3. S. Johnson,
    4. D. Bhattacharya,
    5. M. S. Diamond
    . 2011. Memory B cells, but not long-lived plasma cells, possess antigen specificities for viral escape mutants. J. Exp. Med. 208: 2599–2606.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Pape, K. A.,
    2. J. J. Taylor,
    3. R. W. Maul,
    4. P. J. Gearhart,
    5. M. K. Jenkins
    . 2011. Different B cell populations mediate early and late memory during an endogenous immune response. Science 331: 1203–1207.
    OpenUrlAbstract/FREE Full Text
    1. Dogan, I.,
    2. B. Bertocci,
    3. V. Vilmont,
    4. F. Delbos,
    5. J. Mégret,
    6. S. Storck,
    7. C. A. Reynaud,
    8. J. C. Weill
    . 2009. Multiple layers of B cell memory with different effector functions. Nat. Immunol. 10: 1292–1299.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Kurosaki, T.,
    2. Y. Aiba,
    3. K. Kometani,
    4. S. Moriyama,
    5. Y. Takahashi
    . 2010. Unique properties of memory B cells of different isotypes. Immunol. Rev. 237: 104–116.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Zuccarino-Catania, G. V.,
    2. S. Sadanand,
    3. F. J. Weisel,
    4. M. M. Tomayko,
    5. H. Meng,
    6. S. H. Kleinstein,
    7. K. L. Good-Jacobson,
    8. M. J. Shlomchik
    . 2014. CD80 and PD-L2 define functionally distinct memory B cell subsets that are independent of antibody isotype. Nat. Immunol. 15: 631–637.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Tomayko, M. M.,
    2. N. C. Steinel,
    3. S. M. Anderson,
    4. M. J. Shlomchik
    . 2010. Cutting edge: hierarchy of maturity of murine memory B cell subsets. J. Immunol. 185: 7146–7150.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Klein, U.,
    2. K. Rajewsky,
    3. R. Küppers
    . 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.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Tangye, S. G.,
    2. Y. J. Liu,
    3. G. Aversa,
    4. J. H. Phillips,
    5. J. E. de Vries
    . 1998. Identification of functional human splenic memory B cells by expression of CD148 and CD27. J. Exp. Med. 188: 1691–1703.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Wei, C.,
    2. J. Anolik,
    3. A. Cappione,
    4. B. Zheng,
    5. A. Pugh-Bernard,
    6. J. Brooks,
    7. E. H. Lee,
    8. E. C. Milner,
    9. I. Sanz
    . 2007. A new population of cells lacking expression of CD27 represents a notable component of the B cell memory compartment in systemic lupus erythematosus. J. Immunol. 178: 6624–6633.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Ehrhardt, G. R.,
    2. J. T. Hsu,
    3. L. Gartland,
    4. C. M. Leu,
    5. S. Zhang,
    6. R. S. Davis,
    7. M. D. Cooper
    . 2005. Expression of the immunoregulatory molecule FcRH4 defines a distinctive tissue-based population of memory B cells. J. Exp. Med. 202: 783–791.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Fecteau, J. F.,
    2. G. Côté,
    3. S. Néron
    . 2006. A new memory CD27-IgG+ B cell population in peripheral blood expressing VH genes with low frequency of somatic mutation. J. Immunol. 177: 3728–3736.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Ehrhardt, G. R.,
    2. C. M. Leu,
    3. S. Zhang,
    4. G. Aksu,
    5. T. Jackson,
    6. C. Haga,
    7. J. T. Hsu,
    8. D. M. Schreeder,
    9. R. S. Davis,
    10. M. D. Cooper
    . 2007. Fc receptor-like proteins (FCRL): immunomodulators of B cell function. Adv. Exp. Med. Biol. 596: 155–162.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Ehrhardt, G. R.,
    2. A. Hijikata,
    3. H. Kitamura,
    4. O. Ohara,
    5. J. Y. Wang,
    6. M. D. Cooper
    . 2008. Discriminating gene expression profiles of memory B cell subpopulations. J. Exp. Med. 205: 1807–1817.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Polson, A. G.,
    2. B. Zheng,
    3. K. Elkins,
    4. W. Chang,
    5. C. Du,
    6. P. Dowd,
    7. L. Yen,
    8. C. Tan,
    9. J. A. Hongo,
    10. H. Koeppen,
    11. A. Ebens
    . 2006. Expression pattern of the human FcRH/IRTA receptors in normal tissue and in B-chronic lymphocytic leukemia. Int. Immunol. 18: 1363–1373.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Wilson, T. J.,
    2. A. Fuchs,
    3. M. Colonna
    . 2012. Cutting edge: human FcRL4 and FcRL5 are receptors for IgA and IgG. J. Immunol. 188: 4741–4745.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Ehrhardt, G. R.,
    2. R. S. Davis,
    3. J. T. Hsu,
    4. C. M. Leu,
    5. A. Ehrhardt,
    6. M. D. Cooper
    . 2003. The inhibitory potential of Fc receptor homolog 4 on memory B cells. Proc. Natl. Acad. Sci. USA 100: 13489–13494.
    OpenUrlAbstract/FREE Full Text
    1. Liu, Y.,
    2. K. Bezverbnaya,
    3. T. Zhao,
    4. M. J. Parsons,
    5. M. Shi,
    6. B. Treanor,
    7. G. R. Ehrhardt
    . 2015. Involvement of the HCK and FGR src-family kinases in FCRL4-mediated immune regulation. J. Immunol. 194: 5851–5860.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Sohn, H. W.,
    2. P. D. Krueger,
    3. R. S. Davis,
    4. S. K. Pierce
    . 2011. FcRL4 acts as an adaptive to innate molecular switch dampening BCR signaling and enhancing TLR signaling. Blood 118: 6332–6341.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Weiss, G. E.,
    2. P. D. Crompton,
    3. S. Li,
    4. L. A. Walsh,
    5. S. Moir,
    6. B. Traore,
    7. K. Kayentao,
    8. A. Ongoiba,
    9. O. K. Doumbo,
    10. S. K. Pierce
    . 2009. Atypical memory B cells are greatly expanded in individuals living in a malaria-endemic area. J. Immunol. 183: 2176–2182.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Moir, S.,
    2. J. Ho,
    3. A. Malaspina,
    4. W. Wang,
    5. A. C. DiPoto,
    6. M. A. O’Shea,
    7. G. Roby,
    8. S. Kottilil,
    9. J. Arthos,
    10. M. A. Proschan, et al
    . 2008. Evidence for HIV-associated B cell exhaustion in a dysfunctional memory B cell compartment in HIV-infected viremic individuals. J. Exp. Med. 205: 1797–1805.
    OpenUrlAbstract/FREE Full Text
    1. Kardava, L.,
    2. S. Moir,
    3. W. Wang,
    4. J. Ho,
    5. C. M. Buckner,
    6. J. G. Posada,
    7. M. A. O’Shea,
    8. G. Roby,
    9. J. Chen,
    10. H. W. Sohn, et al
    . 2011. Attenuation of HIV-associated human B cell exhaustion by siRNA downregulation of inhibitory receptors. J. Clin. Invest. 121: 2614–2624.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Jelicic, K.,
    2. R. Cimbro,
    3. F. Nawaz,
    4. D. W. Huang,
    5. X. Zheng,
    6. J. Yang,
    7. R. A. Lempicki,
    8. M. Pascuccio,
    9. D. Van Ryk,
    10. C. Schwing, et al
    . 2013. The HIV-1 envelope protein gp120 impairs B cell proliferation by inducing TGF-β1 production and FcRL4 expression. Nat. Immunol. 14: 1256–1265.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Siewe, B.,
    2. A. J. Nipper,
    3. H. Sohn,
    4. J. T. Stapleton,
    5. A. Landay
    . 2017. FcRL4 expression identifies a pro-inflammatory B cell subset in viremic HIV-infected subjects. Front. Immunol. 8: 1339.
    OpenUrl
  28. ↵
    1. Amara, K.,
    2. E. Clay,
    3. L. Yeo,
    4. D. Ramsköld,
    5. J. Spengler,
    6. N. Sippl,
    7. J. A. Cameron,
    8. L. Israelsson,
    9. P. J. Titcombe,
    10. C. Grönwall, et al
    . 2017. B cells expressing the IgA receptor FcRL4 participate in the autoimmune response in patients with rheumatoid arthritis. J. Autoimmun. 81: 34–43.
    OpenUrl
  29. ↵
    1. Smith, K.,
    2. L. Garman,
    3. J. Wrammert,
    4. N. Y. Zheng,
    5. J. D. Capra,
    6. R. Ahmed,
    7. P. C. Wilson
    . 2009. Rapid generation of fully human monoclonal antibodies specific to a vaccinating antigen. Nat. Protoc. 4: 372–384.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Brochet, X.,
    2. M. P. Lefranc,
    3. V. Giudicelli
    . 2008. IMGT/V-QUEST: the highly customized and integrated system for IG and TR standardized V-J and V-D-J sequence analysis. Nucleic Acids Res. 36: W503–W508.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Godbey, W. T.,
    2. K. K. Wu,
    3. A. G. Mikos
    . 1999. Poly(ethylenimine) and its role in gene delivery. J. Control. Release 60: 149–160.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Ippolito, G. C.,
    2. K. H. Hoi,
    3. S. T. Reddy,
    4. S. M. Carroll,
    5. X. Ge,
    6. T. Rogosch,
    7. M. Zemlin,
    8. L. D. Shultz,
    9. A. D. Ellington,
    10. C. L. Vandenberg,
    11. G. Georgiou
    . 2012. Antibody repertoires in humanized NOD-scid-IL2Rγ(null) mice and human B cells reveals human-like diversification and tolerance checkpoints in the mouse. PLoS One 7: e35497.
    OpenUrlCrossRefPubMed
  33. ↵
    1. McDaniel, J. R.,
    2. B. J. DeKosky,
    3. H. Tanno,
    4. A. D. Ellington,
    5. G. Georgiou
    . 2016. Ultra-high-throughput sequencing of the immune receptor repertoire from millions of lymphocytes. Nat. Protoc. 11: 429–442.
    OpenUrlCrossRefPubMed
  34. ↵
    1. DeKosky, B. J.,
    2. T. Kojima,
    3. A. Rodin,
    4. W. Charab,
    5. G. C. Ippolito,
    6. A. D. Ellington,
    7. G. Georgiou
    . 2015. In-depth determination and analysis of the human paired heavy- and light-chain antibody repertoire. Nat. Med. 21: 86–91.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Zhang, J.,
    2. K. Kobert,
    3. T. Flouri,
    4. A. Stamatakis
    . 2014. PEAR: a fast and accurate Illumina paired-end reAd mergeR. Bioinformatics 30: 614–620.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Bolger, A. M.,
    2. M. Lohse,
    3. B. Usadel
    . 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30: 2114–2120.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Bolotin, D. A.,
    2. S. Poslavsky,
    3. I. Mitrophanov,
    4. M. Shugay,
    5. I. Z. Mamedov,
    6. E. V. Putintseva,
    7. D. M. Chudakov
    . 2015. MiXCR: software for comprehensive adaptive immunity profiling. Nat. Methods 12: 380–381.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Kyte, J.,
    2. R. F. Doolittle
    . 1982. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157: 105–132.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Schickel, J. N.,
    2. S. Glauzy,
    3. Y. S. Ng,
    4. N. Chamberlain,
    5. C. Massad,
    6. I. Isnardi,
    7. N. Katz,
    8. G. Uzel,
    9. S. M. Holland,
    10. C. Picard, et al
    . 2017. Self-reactive VH4-34–expressing IgG B cells recognize commensal bacteria. [Published erratum appears in 2017 J. Exp. Med. 214: 2161.] J. Exp. Med. 214: 1991–2003.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Benckert, J.,
    2. N. Schmolka,
    3. C. Kreschel,
    4. M. J. Zoller,
    5. A. Sturm,
    6. B. Wiedenmann,
    7. H. Wardemann
    . 2011. The majority of intestinal IgA+ and IgG+ plasmablasts in the human gut are antigen-specific. J. Clin. Invest. 121: 1946–1955.
    OpenUrlCrossRefPubMed
  41. ↵
    1. Petrof, E. O.,
    2. G. B. Gloor,
    3. S. J. Vanner,
    4. S. J. Weese,
    5. D. Carter,
    6. M. C. Daigneault,
    7. E. M. Brown,
    8. K. Schroeter,
    9. E. Allen-Vercoe
    . 2013. Stool substitute transplant therapy for the eradication of Clostridium difficile infection: ‘RePOOPulating’ the gut. Microbiome 1: 3.
    OpenUrlCrossRefPubMed
  42. ↵
    1. Tiller, T.,
    2. M. Tsuiji,
    3. S. Yurasov,
    4. K. Velinzon,
    5. M. C. Nussenzweig,
    6. H. Wardemann
    . 2007. Autoreactivity in human IgG+ memory B cells. Immunity 26: 205–213.
    OpenUrlCrossRefPubMed
  43. ↵
    1. Mietzner, B.,
    2. M. Tsuiji,
    3. J. Scheid,
    4. K. Velinzon,
    5. T. Tiller,
    6. K. Abraham,
    7. J. B. Gonzalez,
    8. V. Pascual,
    9. D. Stichweh,
    10. H. Wardemann,
    11. M. C. Nussenzweig
    . 2008. Autoreactive IgG memory antibodies in patients with systemic lupus erythematosus arise from nonreactive and polyreactive precursors. Proc. Natl. Acad. Sci. USA 105: 9727–9732.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Wardemann, H.,
    2. S. Yurasov,
    3. A. Schaefer,
    4. J. W. Young,
    5. E. Meffre,
    6. M. C. Nussenzweig
    . 2003. Predominant autoantibody production by early human B cell precursors. Science 301: 1374–1377.
    OpenUrlAbstract/FREE Full Text
  45. ↵
    1. Meffre, E.,
    2. A. Louie,
    3. J. Bannock,
    4. L. J. Kim,
    5. J. Ho,
    6. C. C. Frear,
    7. L. Kardava,
    8. W. Wang,
    9. C. M. Buckner,
    10. Y. Wang, et al
    . 2016. Maturational characteristics of HIV-specific antibodies in viremic individuals. JCI Insight 1: e84610.
    OpenUrl
  46. ↵
    1. Zinöcker, S.,
    2. C. E. Schindler,
    3. J. Skinner,
    4. T. Rogosch,
    5. M. Waisberg,
    6. J. N. Schickel,
    7. E. Meffre,
    8. K. Kayentao,
    9. A. Ongoïba,
    10. B. Traoré,
    11. S. K. Pierce
    . 2015. The V gene repertoires of classical and atypical memory B cells in malaria-susceptible West African children. J. Immunol. 194: 929–939.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    1. Lau, D.,
    2. L. Y. Lan,
    3. S. F. Andrews,
    4. C. Henry,
    5. K. T. Rojas,
    6. K. E. Neu,
    7. M. Huang,
    8. Y. Huang,
    9. B. DeKosky,
    10. A. E. Palm, et al
    . 2017. Low CD21 expression defines a population of recent germinal center graduates primed for plasma cell differentiation. Sci. Immunol. 2: eaai8153.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Bunker, J. J.,
    2. S. A. Erickson,
    3. T. M. Flynn,
    4. C. Henry,
    5. J. C. Koval,
    6. M. Meisel,
    7. B. Jabri,
    8. D. A. Antonopoulos,
    9. P. C. Wilson,
    10. A. Bendelac
    . 2017. Natural polyreactive IgA antibodies coat the intestinal microbiota. Science 358: eaan6619.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    1. Moir, S.,
    2. A. S. Fauci
    . 2017. B-cell responses to HIV infection. Immunol. Rev. 275: 33–48.
    OpenUrlCrossRef
  50. ↵
    1. Bannard, O.,
    2. J. G. Cyster
    . 2017. Germinal centers: programmed for affinity maturation and antibody diversification. Curr. Opin. Immunol. 45: 21–30.
    OpenUrlCrossRef
  51. ↵
    1. Nowosad, C. R.,
    2. K. M. Spillane,
    3. P. Tolar
    . 2016. Germinal center B cells recognize antigen through a specialized immune synapse architecture. Nat. Immunol. 17: 870–877.
    OpenUrlCrossRef
  52. ↵
    1. Khalil, A. M.,
    2. J. C. Cambier,
    3. M. J. Shlomchik
    . 2012. B cell receptor signal transduction in the GC is short-circuited by high phosphatase activity. Science 336: 1178–1181.
    OpenUrlAbstract/FREE Full Text
  53. ↵
    1. Jourdan, M.,
    2. N. Robert,
    3. M. Cren,
    4. C. Thibaut,
    5. C. Duperray,
    6. A. Kassambara,
    7. M. Cogné,
    8. K. Tarte,
    9. B. Klein,
    10. J. Moreaux
    . 2017. Characterization of human FCRL4-positive B cells. PLoS One 12: e0179793.
    OpenUrl
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section