Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Aug 14;8:910.
doi: 10.3389/fimmu.2017.00910. eCollection 2017.

Quantitative Analysis of Repertoire-Scale Immunoglobulin Properties in Vaccine-Induced B-Cell Responses

Affiliations
Free PMC article

Quantitative Analysis of Repertoire-Scale Immunoglobulin Properties in Vaccine-Induced B-Cell Responses

Ilja V Khavrutskii et al. Front Immunol. .
Free PMC article

Abstract

Recent advances in the next-generation sequencing of B-cell receptors (BCRs) enable the characterization of humoral responses at a repertoire-wide scale and provide the capability for identifying unique features of immune repertoires in response to disease, vaccination, or infection. Immunosequencing now readily generates 103-105 sequences per sample; however, statistical analysis of these repertoires is challenging because of the high genetic diversity of BCRs and the elaborate clonal relationships among them. To date, most immunosequencing analyses have focused on reporting qualitative trends in immunoglobulin (Ig) properties, such as usage or somatic hypermutation (SHM) percentage of the Ig heavy chain variable (IGHV) gene segment family, and on reducing complex Ig property distributions to simple summary statistics. However, because Ig properties are typically not normally distributed, any approach that fails to assess the distribution as a whole may be inadequate in (1) properly assessing the statistical significance of repertoire differences, (2) identifying how two repertoires differ, and (3) determining appropriate confidence intervals for assessing the size of the differences and their potential biological relevance. To address these issues, we have developed a technique that uses Wilcox' robust statistics toolbox to identify statistically significant vaccine-specific differences between Ig repertoire properties. The advantage of this technique is that it can determine not only whether but also where the distributions differ, even when the Ig repertoire properties are non-normally distributed. We used this technique to characterize murine germinal center (GC) B-cell repertoires in response to a complex Ebola virus-like particle (eVLP) vaccine candidate with known protective efficacy. The eVLP-mediated GC B-cell responses were highly diverse, consisting of thousands of clonotypes. Despite this staggering diversity, we identified statistically significant differences between non-immunized, vaccine only, and vaccine-plus-adjuvant groups in terms of Ig properties, including IGHV-family usage, SHM percentage, and characteristics of the BCR complementarity-determining region. Most notably, our analyses identified a robust eVLP-specific feature-enhanced IGHV8-family usage in B-cell repertoires. These findings demonstrate the utility of our technique in identifying statistically significant BCR repertoire differences following vaccination. More generally, our approach is potentially applicable to a wide range of studies in infection, vaccination, auto-immunity, and cancer.

Keywords: B cell; Ebola; clonotype; germinal center; immunoglobulin; immunosequencing; repertoire properties; statistical analysis.

Figures

Figure 1
Figure 1
Overview of the approach.
Figure 2
Figure 2
Ebola virus-like particle (eVLP)-mediated germinal center (GC) reactions following a single IM vaccination. Mice were vaccinated (intramuscularly via the caudal muscle of the right hind leg) with either eVLP alone or eVLP plus poly-ICLC (pICLC), and draining lymph nodes were isolated at the indicated time points. Single-cell suspensions were stained with B220, IgD, IgM, CD38, CD95, GL-7, and live/dead dye, after which they were collected by FACS. (A) Representative FACS plots of day 10 (D10) and D21 B220+CD95+GL7+ GC B cells. (B) Relative frequency following vaccination; error bars = SEM; *p < 0.05; n = 5 per group. (C) Serum samples were collected at the indicated time points and IgG responses against EBOV-GP1,2 following vaccination were measured by enzyme-linked immunosorbent assay. Endpoint titers were determined at a serum dilution corresponding to +0.2 SD from control OD readings (***p < 0.001; n = 3 per group).
Figure 3
Figure 3
Clonotype-invariant Ig heavy chain variable-family usage. (A) The stacked bar charts show individual subject contributions. Subjects are numbered S1 through S8 in decreasing order of clonotype number within each group. Comparison of (B) Ebola virus-like particle (eVLP) and (C) eVLP/poly-ICLC (pICLC) groups with control group. Statistically significant differences are shown in the difference plots of (B,C) with open circles. The shaded areas around the main solid curves show confidence intervals.
Figure 4
Figure 4
Clonotype-weighted SHM percentage in Ig heavy chain variable-gene segment (V-SHM). (A) The stacked bar charts show individual subject contributions. Subjects are numbered S1 through S8 in decreasing order of clonotype number within each group. Comparison of (B) Ebola virus-like particle (eVLP) and (C) eVLP/poly-ICLC (pICLC) groups with the control group. Statistically significant differences are shown in the difference plots of (B,C) with open circles. Although barely noticeable, the shaded areas around the main solid curves show confidence intervals.
Figure 5
Figure 5
Clonotype-invariant complementarity-determining region (CDR3) length. (A) The stacked bar charts show individual subject contributions. Subjects are numbered S1 through S3 in decreasing order of clonotype number within each group. Comparison of (B) Ebola virus-like particle (eVLP) and (C) eVLP/poly-ICLC (pICLC) groups with the control group. The shaded areas around the main solid curves show confidence intervals. See text for definitions of all clonotypes (AC), branched clonotypes (BC), and solitary clonotypes (SC) repertoire subsets.
Figure 6
Figure 6
Direct comparison of Ebola virus-like particle (eVLP) and eVLP/poly-ICLC (pICLC groups). (A) Clonotype-invariant Ig heavy chain variable-family usage. (B) Clonotype-weighted SHM percentage in V-gene segment (V-SHM). (C) Clonotype-invariant complementarity-determining region (CDR3) length. Open circles indicate statistically significant differences in the difference plots. The shaded areas around the main solid curves show confidence intervals.

Similar articles

See all similar articles

Cited by 3 articles

References

    1. Cobey S, Wilson P, Matsen FA. The evolution within us. Philos Trans R Soc Lond B Biol Sci (2015) 370(1676):20140235.10.1098/rstb.2014.0235 - DOI - PMC - PubMed
    1. Hoehn KB, Fowler A, Lunter G, Pybus OG. The diversity and molecular evolution of B-cell receptors during infection. Mol Biol Evol (2016) 33(5):1147–57.10.1093/molbev/msw015 - DOI - PMC - PubMed
    1. Greiff V, Bhat P, Cook SC, Menzel U, Kang W, Reddy ST. A bioinformatic framework for immune repertoire diversity profiling enables detection of immunological status. Genome Med (2015) 7(1):49.10.1186/s13073-015-0169-8 - DOI - PMC - PubMed
    1. Fairfax KA, Kallies A, Nutt SL, Tarlinton DM. Plasma cell development: from B-cell subsets to long-term survival niches. Semin Immunol (2008) 20(1):49–58.10.1016/j.smim.2007.12.002 - DOI - PubMed
    1. Kuraoka M, Schmidt AG, Nojima T, Feng F, Watanabe A, Kitamura D, et al. Complex antigens drive permissive clonal selection in germinal centers. Immunity (2016) 44(3):542–52.10.1016/j.immuni.2016.02.010 - DOI - PMC - PubMed
Feedback