High-throughput sequencing of the zebrafish antibody repertoire

Abstract

Despite tremendous progress in understanding the nature of the immune system, the full diversity of an organism's antibody repertoire is unknown. We used high-throughput sequencing of the variable domain of the antibody heavy chain from 14 zebrafish to analyze VDJ usage and antibody sequence. Zebrafish were found to use between 50 and 86% of all possible VDJ combinations and shared a similar frequency distribution, with some correlation of VDJ patterns between individuals. Zebrafish antibodies retained a few thousand unique heavy chains that also exhibited a shared frequency distribution. We found evidence of convergence, in which different individuals made the same antibody. This approach provides insight into the breadth of the expressed antibody repertoire and immunological diversity at the level of an individual organism.

Fig. 1

(A) Schematic drawing of the VDJ recombination of an antibody heavy-chain gene, the cDNA amplicon library construction, and the infomatics pipeline. The heavy-chain VDJ segment of an antibody is created by recombination, junctional diversity, and hypermutation. We designed primer sets to amplify the expressed heavy-chain mRNA, which were then sequenced and analyzed as outlined. High-throughput sequencing allows determination of the identity of nearly all heavy-chain sequences. (B) Gender and family information for the 14 sequenced zebrafish.

Fig. 2

The entire expressed VDJ repertoires for individual fish g, h, j, and k (top to bottom). The three axes enumerate all possible V, D, and J values, so each point in three-space is a unique VDJ combination. Both the size of the sphere at each point and the intensity correspond to the number of reads matching that particular VDJ combination. Gray scale is plotted on a linear scale, and the dot size is plotted on a log scale. The upper limits of the scales are set to the most populated VDJ combination for each fish, with PCR bias factored out.

Fig. 3

VDJ repertoire analysis for all 14 fish. (A) Abundance distribution for each VDJ combination. A small number of VDJ combinations are highly represented in each fish, and most VDJ combinations are represented only at low abundance. The shape of the distribution is common among all of the fish sampled. This histogram is oriented sideways (from left to right) to emphasize that a small number of VDJ combinations are highly abundant, with a distribution that falls off rapidly. (B) Rarefaction analysis of VDJ diversity demonstrates that as one sequences more deeply into a fish, the number of new VDJ classes discovered saturates. (C) Histogram of correlations between VDJ repertoires. The data are collected as histograms and compared to simulated fish which have random VDJ repertoires. The simulated fish have no significant correlations, whereas some of the real fish have high correlations, representing 5 SD outliers of the random model. The highest correlations are from males in the same family (table S5A). (D) When the largest VDJ class in each fish is eliminated, the correlations are reduced and there is a larger proportion of moderate female correlations.

Fig. 4

Antibody heavy-chain repertoire diversity estimates for IgM in all 14 fish. (A) Rarefaction analysis of heavy-chain diversity demonstrates that as one sequences more deeply into a fish, the number of new antibodies discovered saturates at afew thousand. (B) Antibody abundance distributions for each fish. This histogram is oriented sideways (from left to right) to emphasize that a small number of antibodies (clusters) are highly abundant, with a distribution that falls off rapidly as a power law. The shape of the distribution is universal among all of the fish sampled. The bend at small abundance is caused by variability in the total reads sampled per fish and is not significant. (C) Total antibody diversity estimates for IgM using different criteria. VDJ diversity is the number of VDJ classes per fish, as described in Fig. 3A. Antibodies observed (>2 reads; VDJ classes composed only of antibody clusters with two or fewer reads are counted as one) is the number of unique antibodies per fish described in Fig. 4A. Capture-recapture estimate 1 refers to an estimate based on observed antibody abundances (14). Capture-recapture estimate 2 refers to an estimate using equal probability of all antibodies. Antibodies observed, undercount corrected refers to the upper bound. (D) Histogram of number of fish with shared IgM sequences. Hundreds of sequences are shared between pairs of fish, while a few tens of sequences are shared between three fish. Five sequences are shared between four or more fish, and none are shared among all fourteen fish. Sequence comparisons without mutations incorporate differences at the V/D and D/J junctions alone. Convergence on the amino acid level is also plotted.