The basis for haplotype complexity in VCBPs, an immune-type receptor in amphioxus

Abstract

Innate immune gene repertoires are restricted primarily to germline variation. Adaptive immunity, by comparison, relies on somatic variation of germline-encoded genes to generate extraordinarily large numbers of non-heritable antigen recognition motifs. Invertebrates lack the key features of vertebrate adaptive immunity, but have evolved a variety of alternative mechanisms to successfully protect the integrity of "self"; in many cases, these appear to be taxon-specific innovations. In the protochordate Branchiostoma floridae (amphioxus), the variable region-containing chitin-binding proteins (VCBPs) constitute a multigene family (comprised of VCBPs 1-5), which possesses features that are consistent with innate immune-type function. A large number of VCBP alleles and haplotypes are shown to exhibit levels of polymorphism exceeding the elevated overall levels determined for the whole amphioxus genome (JGI). VCBP genes of the 2 and 5 types are distinguished further by a highly polymorphic segment (exon 2) in the N-terminal immunoglobulin domain, defined previously as a "hypervariable region" or a "hotspot." Genomic deoxyribonucleic acid (DNA) and complementary DNA (cDNA) sequences from large numbers of animals representing different populations reveal further significant differences in sequence complexity within and across VCBP2/5 haplotypes that arise through overlapping mechanisms of genetic exchange, gene copy number variation as well as mutation and give rise to distinct allelic lineages. The collective observations suggest that mechanisms were in place at the time of divergence of the cephalochordates that could selectively hyperdiversify immune-type receptors within a multigene family.

Fig. 1

Genomic organization of the VCBP2/5 gene cluster. A In Brafl1, the reference genome animal, two VCBP5 paralogs are separated by a VCBP2 gene, which lies within a larger chromosomal region encoding other VCBP genes (Dishaw et al. 2008). Chromosomal distances in (A) are to scale. B Copy number variants (CNVs) observed in different VCBP2/5 haplotypes are depicted in grey shading. C VCBP2 and VCBP5 genes consist of seven exons (EX1-7) that encode two V-type Ig domains and one chitin-binding domain as illustrated for the 5b gene locus. A region of elevated, localized, polymorphism in exon 2 (EX2, hotspot motif) is shaded. Nucleotide sequence length in the reference BAC contig 63n5-43b24 is indicated below introns and exons. Arrows below EX1 and EX3 depict the location of priming sites that were used for PCR genotyping of 5a, 2b, and 5b genes

Fig. 2

Evidence for recombination involving different 5a–2b–5b haplotypes. A The representative EX2 hotspot motifs for each locus are compared in reference haplotypes: BAC clone 62d19, BAC contig 63n5-43b24, and PAC clone 37d15. Haplotypes recovered from animal LP08 and BAC 63n5 are conserved. Haplotype 37d15 shares homology with LP08 and 63n5 across the 2b and 5b loci, but not at the 5a locus. B PAC 37d15 possesses a segment of an upstream VCBP2-like CNV that is interrupted by cloning. The 5a allele of 37d15 shows evidence for mosaicism with a 5b-type allele and is designated 5a/b. A related sequence flanking 2b-type alleles is evident upstream of the 5a (5a/b) locus in 37d15. The 37d15 haplotype may have originated through unequal crossover between 5a and 5b genes. Rectangular boxes in (B) imply full-length gene loci

Fig. 3

Exon 2 motif type and subtype variation generates diversified allelic lineages. A Comparison of two subtypes of -SAG-type 2b alleles (-SAG-motif in bold; defined in mixed animal surveys are distinguished by grey shading). Polymorphisms representing point mutations (reverse image) account for additional variation. B Alleles from additional animals exhibit subtype variation (reverse image) from (A). C Shared polymorphisms in EX2 (boxed) extend to adjacent introns (e.g., SW01 and LP07) or intron variation in otherwise conserved alleles (e.g., LP02 and PACV). Sequence differences can be extensive (e.g., LP10). D Schematic illustration of EX2 polymorphism and relatedness; EX2 types and subtype variants give rise to diverse allelic lineages. In this example, three types of VCBP2 alleles [defined by core EX2 peptide motifs (i.e., “types”): DREY, SAG and SASY] can be classified into subtypes, exhibiting further polymorphic variation (terminal branches; subtype variants) that can result in additional subtype lineages (e.g., NAG and SAAAG). Some VCBP2 alleles (i.e., alternative EX2 types) represent paralogous genes from haplotype-specific CNVs. Bold lines for -NAG- and -SAAG- imply novel lineages of allele types. PacV=Bf Pac 47J9. Numbers in (A) refer to individual sequence clones. Figure in (C) is not meant to depict phylogenetic relationships

Fig. 4

VCBP2 genes with VCBP5b-type EX2 motifs; exceptions to the EX2 allele-type classification rules confound accurate placement of VCBP2/5-type alleles. A 5b-type EX2 polypeptide sequences exhibit subtype variation (reverse image). B VCBP2 cDNAs (shaded, dark grey) have been identified with similar EX2 peptides. Conserved VCBP5b-type EX2 motif is shaded (light grey), intended for reference purposes only

Fig. 5

A hypothetical evolutionary scenario for VCBP2/5 haplotype diversification via sequence change: (1) duplication generated at least three VCBP2/5 genes, each consisting of seven exons; (2) differentiation of gene loci; (3) exchange has occurred between genes 5a and 5b in EX1–2 as well as in EX5–7. EX1–EX3 have undergone genetic exchange between genes 2b and 5b. EX1 of the 5b gene was replaced by unrelated or distantly related genes. All or part of EX7 was exchanged between 5a and 5b. The VCBP 5b locus is used as a reference (except for the unusual EX1), likely owing to more functional constraint resulting in less polymorphism at this locus