An evolutionarily mobile antigen receptor variable region gene: doubly rearranging NAR-TcR genes in sharks

Abstract

Distinctive Ig and T cell receptor (TcR) chains define the two major lineages of vertebrate lymphocyte yet similarly recognize antigen with a single, membrane-distal variable (V) domain. Here we describe the first antigen receptor chain that employs two V domains, which are generated by separate VDJ gene rearrangement events. These molecules have specialized "supportive" TcRdeltaV domains membrane-proximal to domains with most similarity to IgNAR V. The ancestral NAR V gene encoding this domain is hypothesized to have recombined with the TRD locus in a cartilaginous fish ancestor >200 million years ago and encodes the first V domain shown to be used in both Igs and TcRs. Furthermore, these data support the view that gamma/delta TcRs have for long used structural conformations recognizing free antigen.

Fig. 1.

NAR-TcRV is a V domain supported on TcRδ. (A) Products of 5′ RACE PCR with TRDC primers using RNA from shark PBL, spleen, and thymus. (B) Cartoon depiction of IgNAR and the hypothetical γ/δ TcR with NAR-TcR antigen receptors. Stars mark CDR3s generated from somatic rearrangement. (C) Predicted amino acid sequences of four families of NAR-TcRV aligned with IgNARV from nurse shark (nsNART1, U18721; nsNART2, U18680), wobbeygong (wgNART2a, AF336092; wgNART2b, AF336091), and guitarfish (gfNAR, AY524298). Residues conserved are highlighted in gray, and Ig superfamily canonical intradomain and putative interdomain cysteines are highlighted in black. β-Strands and CDRs are noted above the alignment. Percent amino acid identity to the top sequence is shown on the right. (D) Predicted amino acid sequences of eight families of TcRδV, four of which support NAR-TcRV. Highlighting and notation are the same as in C.

Fig. 2.

Phylogenetic analysis of Ig superfamily V domains. See Table 1 and Fig. 6, which are published as supporting information on the PNAS web site, for accession numbers of sequences used as well as the same data presented in a radial phylogram.

Fig. 3.

NAR-TcR is expressed in shark lymphoid tissues and is found in modern sharks. (A) Tissues used in the Northern blot are muscle, thymus, liver, PBLs, gill, spiral valve (intestine), stomach, spleen, pancreas, epigonal (a primary lymphoid organ of cartilaginous fishes), ovary, thymocytes, and brain. Marker (in kb) is shown flanking the blots. Probes were generated from V domain encoding genes of NTV and IgNAR, TRDC, and the loading control NDPK. An arrow marks the cross-hybridizing IgNAR band, and asterisks highlight the NAR-TcR encoding transcripts recognized by TRDC probing a blot allowed to migrate further. (B) Southern blot of HindIII- and PstI-digested genomic DNA probed with NTV1 and IgNAR 7A. Marker (in kb) is shown on left, and higher or lower wash stringency is indicated.

Fig. 4.

NAR-TcR is doubly rearranging. (A) Focus on amino acid sequences of both NAR-TcRV and TcRδV CDR3s. Bold abbreviations indicate NAR-TcRV, TcRδV, and TcRδJ (unpublished data) families used in clones named at left. Conserved cysteine of the f strand and GXG motif of the g strand are highlighted flanking each CDR3. Known and predicted D-segment-encoded amino acids are underlined, “∼” are gaps introduced to align Vs and Js, and “/” marks frameshifts resulting from nonfunctional rearrangements. (B) Diagrams of clones showing NTV assembly mechanism. Recombination and splicing signals are marked by filled triangles for RSSs and open squares for GT/AG intronic splice sites. An incompletely rearranged 5′ RACE clone from PBL shows RSS 3′ of TRDD and 5′ of TRDJ. These data suggest that the mammalian RSS system for δ is also used in shark (Inset) (52, 53).

Fig. 5.

Hypothetical scheme for history of NAR variable domains. Depictions of genomic loci are simplified; filled bars indicate variable segments with leader exons upstream. Direction of cross-hatching distinguishes TRDV segments dedicated to supporting NAR-TcRV.