Evolutionarily conserved TCR binding sites, identification of T cells in primary lymphoid tissues, and surprising trans-rearrangements in nurse shark

Abstract

Cartilaginous fish are the oldest animals that generate RAG-based Ag receptor diversity. We have analyzed the genes and expressed transcripts of the four TCR chains for the first time in a cartilaginous fish, the nurse shark (Ginglymostoma cirratum). Northern blotting found TCR mRNA expression predominantly in lymphoid and mucosal tissues. Southern blotting suggested translocon-type loci encoding all four chains. Based on diversity of V and J segments, the expressed combinatorial diversity for gamma is similar to that of human, alpha and beta may be slightly lower, and delta diversity is the highest of any organism studied to date. Nurse shark TCRdelta have long CDR3 loops compared with the other three chains, creating binding site topologies comparable to those of mammalian TCR in basic paratope structure; additionally, nurse shark TCRdelta CDR3 are more similar to IgH CDR3 in length and heterogeneity than to other TCR chains. Most interestingly, several cDNAs were isolated that contained IgM or IgW V segments rearranged to other gene segments of TCRdelta and alpha. Finally, in situ hybridization experiments demonstrate a conservation of both alpha/beta and gamma/delta T cell localization in the thymus across 450 million years of vertebrate evolution, with gamma/delta TCR expression especially high in the subcapsular region. Collectively, these data make the first cellular identification of TCR-expressing lymphocytes in a cartilaginous fish.

FIGURE 1

Nurse shark TCR V gene segment amino acid alignments. TCRα (A), TCRβ (B), TCRγ (C), and TCRδ (D). Conserved residues of Ag receptor variable domains are marked above the first sequence of each set (36). Gray highlighting marks sites of potential N-linked glycosylation (NxS/T, where x is any residue other than proline). Dashes indicate gaps in alignment, and dots mark identity to the first sequence. Percent amino acid identity compared with the first sequence is shown on the right. Predicted β strands are indicated below each alignment (–39). Highlighted δVs are found only supporting N-terminal NAR-TCRV and therefore do not have leader sequences immediately preceding (15). Four spaces separate predicted leader peptide from mature protein. V sequences preceded by an asterisk (*) were only isolated with absent (unspliced) leader peptide exons, whereas those preceded by a number sign (#) are known only from incomplete 5′ cDNA clones. For contrast, the hagfish APAR-A sequence is shown aligned at the bottom (accession number BAD90578, www.ncbi.nlm.nih.gov/sites/entrez).

FIGURE 2

Phylogenetic analysis of cartilaginous fish TCR βV (A) and δV gene (B) families. Neighbor-joining tree with genetic distance scale bar shown. Numbers at nodes represent percentage support from 1000 bootstrap replications. NAR-TCR supporting δV are outlined in bold. Species: H. francisci (horn shark), Ginglymostoma cirratum (nurse shark), R. eglanteria (clearnose skate), and elephant shark (C. milii) (46).

FIGURE 3

Trans rearrangements between Ig V domains and TCRα and δ gene components. A, Amino acid alignment of clones; slashes indicate frame shifts in CDR3. Gene segments contributing are listed at the top of alignment. Gray highlighted C sequence is from TCRα, and C genes from other five clones are TCRδ. B, Drawing showing derivation of gene segments in these six chimeric transcripts. Top panel, First five sequences in A. Bottom panel, TCRα clone 21A1.

FIGURE 4

Amino acid alignment of TCR C domains of representative vertebrates. TCRα (A), TCRβ (B), TCRγ (C), and TCRδ (D). The predicted IgSF domain, transmembrane region, and cytoplasmic tail are indicated above the alignments and the β strands beneath. Dashes mark gaps introduced into alignment. Canonical cysteines that make intradomain and interchain disulfide bonds are highlighted in black, as are putative N-linked glycosylation sites. More or less conserved residues are indicated by dark and light gray highlighting, respectively. Accession numbers for other species are as follows: horned shark: β AAA61563, δ AAA87016; skate: α AAB51495, β AAB51496, γ AAB51498, δ AAB51497; flounder: α BAC65457, β BAC65459, γ BAC65461, δ BAC65464; axolotl: α AAA98473, β AAA48534, δ AY029365; lamprey: TCR-like AAU09668; frog: β BAC67174, γ AAM21541; chicken: α AAC60277, β AB092341, γ AAA87009, δ AAD51740; cow: α AAO42514, β BAA14168, δ BC104586 (predicted from nucleotide); mouse: α AAB47020, β DQ340294, γ CAA25294, δ AAA51274; and human: α AAO72258, β AAO72258, γ M16768, δ A31326.

FIGURE 5

Genomic Southern and Northern blotting. Probes to the C domain Ig regions of each TCR chain were used to probe blotted nucleic acid agarose gels. A, Genomic DNA of three individual nurse sharks (a, b, c) digested (from left to right) with BamHI, EcoRI, HindIII, PstI, and Sac I. B, RNA from various nurse shark tissues. Size of migration markers is shown in kilobases.

FIGURE 6

In situ hybridization of shark thymus. Positive hybridization is purple. TCRαC probe (A, B), TCRβC probe (C, D), TCRγC probe (E, F), TCRδC probe (G, H), MHC class Ia probe (I, J), MHC class IIb probe (K, L), RAG1 probe (M, N), and TdT probe (O, P). Scale bar, 100 μM; original magnification ×10 and ×20. Negative controls with sense probes (not shown) showed no staining. c, cortex; m, medulla; t, trabeculae.