Conservation of Toll-Like Receptor Signaling Pathways in Teleost Fish

Abstract

In mammals, Toll-like receptors (TLR) recognize ligands, including pathogen-associated molecular patterns (PAMPs), and respond with ligand-specific induction of genes. In this study, we establish evolutionary conservation in teleost fish of key components of the TLR-signaling pathway that act as switches for differential gene induction, including MYD88, TIRAP, TRIF, TRAF6, IRF3, and IRF7. We further explore this conservation with a molecular phylogenetic analysis of MYD88. To the extent that current genomic analysis can establish, each vertebrate has one ortholog to each of these genes. For molecular tree construction and phylogeny inference, we demonstrate a methodology for including genes with only partial primary sequences without disrupting the topology provided by the high-confidence full-length sequences. Conservation of the TLR-signaling molecules suggests that the basic program of gene regulation by the TLR-signaling pathway is conserved across vertebrates. To test this hypothesis, leukocytes from a model fish, rainbow trout (Oncorhynchus mykiss), were stimulated with known mammalian TLR agonists including: diacylated and triacylated forms of lipoprotein, flagellin, two forms of LPS, synthetic double-stranded RNA, and two imidazoquinoline compounds (loxoribine and R848). Trout leukocytes responded in vitro to a number of these agonists with distinct patterns of cytokine expression that correspond to mammalian responses. Our results support the key prediction from our phylogenetic analyses that strong selective pressure of pathogenic microbes has preserved both TLR recognition and signaling functions during vertebrate evolution.

Figure 1

A cartoon of TLR signaling. This simplified cartoon illustrates the role key conserved signaling molecules might play in adapting the input from TLR agonists to produce conserved induction of specific batteries of genes. The topology of this cartoon is greatly simplified to facilitate discussion. For example, TLR3, TLR7, TLR8, and TLR9 typically interact with their ligands in endosomes. Citations for cartoons that more completely represent the current body of knowledge are provided in the text.

Figure 2

Molecular tree of MYD88. Sequences in black are full-length high-confidence gene predictions; branches in black form the core tree predicted from these sequences. Sequences in grey are partial sequences that have been placed without altering the core tree. Grey branch lengths are influenced by the length of the partial sequence.

Figure 3

(not to scale). Order and orientation of genes syntenic to MYD88 in sequenced genomes. Since genomes are unfinished, a small possibility exits that any gene portrayed as absent is actually present. Orthologs are identified by the HUGO symbol of the human ortholog, with the exception of “DASH”, which identifies genes orthologous to the DASH cryptochrome (GeneID 402986). MYD88 is always followed by OSR1 and SLC22A13. MYD88 is preceded by genes that vary slightly on a theme most completely represented in the Xenopus genome. Genes are intentionally aligned in columns to facilitate visualization of synteny. Such alignments help guide the search for conserved regulatory elements. The methodology of identifying orthologs is indicated for each genome, as well as the approximate length in kilobases of the region portrayed. Order and orientation of genes syntenic to MYD88 in sequenced genomes. Since genomes are unfinished, a small possibility exits that any gene portrayed as absent is actually present. Orthologs are identified by the HUGO symbol of the human ortholog, with the exception of “DASH”, which identifies genes orthologous to the DASH cryptochrome (GeneID 402986). MYD88 is always followed by OSR1 and SLC22A13. MYD88 is preceded by genes that vary slightly on a theme most completely represented in the Xenopus genome. Genes are intentionally aligned in columns to facilitate visualization of synteny. Such alignments help guide the search for conserved regulatory elements. The methodology of identifying orthologs is indicated for each genome, as well as the approximate length in kilobases of the region portrayed.