Although TLR7 and TLR8 are phylogenetically very close, their natural ligand, ssRNA of viral origin, stimulates human TLR7 and TLR8 and mouse TLR7, but not mouse TLR8 [145, 146]. Although mouse TLR8 is functional, it displays differences with respect to ligand specificity to human TLR8 [49]. For example, human TLR8, but not mouse TLR8, recognizes the imiquimod compound R848 [147]. Anti-viral RNA-like resiquimod R848 and S28463 and imiquimod compounds (R837, S26308) are members of a group of low molecular weight compounds, the imidazoquinolinamines, that have proven to induce antiviral activity via endogenous cytokine production and are often used to study TLR7 and TLR8.
TLR7 has a remarkable conservation across vertebrates [41, 42] with a relatively low evolutionary rate of the LRR domains, suggesting that ligand specificity could be well conserved. Both tlr7 and tlr8 were first identified in pufferfish [66] and zebrafish [51, 57]. Common carp
tlr7 gene expression can be up-regulated in head kidney cells stimulated with imiquimod, after
which these cells produce elevated levels of pro-inflammatory and type I ifn cytokines mRNA [148]. This could indeed point at a conservation of ligand specificity for TLR7. Grass carp tlr7 gene expression was rapidly up-regulated in spleen, but down-regulated in hepatopancreas after infection with GCRV, a dsRNA virus. Furthermore, tlr7 gene expression in C. idella kidney (CIK) cells was up-regulated following stimulation with poly I:C as TLR3 ligand [149]. Grass carp tlr8
6
gene expression in spleen and head kidney was up-regulated at 24 h post-infection with GCRV. In contrast to tlr7, which gene expression was up-regulated in CIK cells following stimulation with poly I:C [149], tlr8 transcription was rapidly down-regulated by poly I:C in a dose and time-dependent manner [150]. Short hairpin-based inhibition of tlr8 gene expression in CIK cells slightly increased
tlr7 basal gene expression. TLR8 knock-down induced a strong resistance against GCRV [150],
suggesting TLR8 might play a negative role in the antiviral immune response of grass carp. Overall, it remains a challenge to dissect clearly, the existing differences in response of TLR7 and TLR8 to poly I:C and other ligands. The high conservation of TLR7 could indicate a conservation of ligand specificity for ssRNA.
Single copy genes for tlr7 and tlr9 have been found in channel catfish, whereas tlr8 has two representatives. Channel catfish TLR7 is 58.1% identical to human TLR7 [42], in support of the remarkable conservation across vertebrates. In comparison, TLR8 sequences have a lower percentage of identity to human TLR8 (45% approximately). The two catfish tlr8 genes are not located on the same chromosome and probably did not result from a recent duplication [151]. Both catfish tlr8 sequences seem to be closer to the tlr8a, rather than tlr8b of zebrafish [42]. Ligand specificity of TLR7 and TLR8 of catfish has not been resolved yet.
Two tlr7/8 loci were identified from a rainbow trout BAC library using DNA fingerprinting and genetic linkage analyses [152]. Trout tlr7 and tlr8 were found in duplicate copies, but one of the TLR7 genes is present as putative pseudogene. Stimulation with R848 and poly I:C produced elevated levels of pro-inflammatory and type I IFN cytokines mRNA in rainbow trout anterior kidney leukocytes. Gene expression of the tlr7 and tlr8a1 genes themselves were not affected by these treatments, but tlr8a2 expression was moderately down-regulated by R848. Inhibition of acidification of the endosome did not clearly modulate R848-induced cytokine expression, however, so it remains questionable whether recognition of R848 in rainbow trout requires endosomal maturation [152].
Early studies in Atlantic salmon described clear effects of typical ligands for endosomal TLRs, including poly I:C (tlr3), imiquimod R837 (tlr7) and CpG-ODN (tlr9) on gene expression of, among others, ifn and mx genes in liver and head kidney. One major difference between gene induction by S27609 (an analog of imiquimod R837) and poly I:C was that S27609 induced much lower levels of type I ifn, possibly because the two ligands were not always sensed by the same cells [153]. tlr8 gene expression of Atlantic salmon was tissue-restricted with a high level of gene expression in the spleen. Although tlr8 gene expression could be up-regulated in TO cells treated with recombinant type I and type II IFN, TLR8, but not MyD88, gene expression in spleen of infected fish was not affected by challenge with salmon alphavirus subtype 3 (ssRNA). In vitro stimulation of salmon head kidney leukocytes with CpG ODNs and type II ifn gamma also up-regulated myd88 gene expression, but not gene expression of tlr8 [154]. Recently, as already discussed above in the context of TLR3, R848 was shown to induce a typical type I IFN response in Atlantic salmon [138]. Further, fluorescence
in situ hybridization showed that poly I:C induced IFNa and IFNc in a variety of cells in several
organs, whereas R848 induced coexpression of IFNb and IFNc in distinct cells in head kidney and spleen. The latter could be specialized high IFN producers. Most recently, studies have identified two tlr7 genes in Atlantic salmon (of which one is possibly a pseudogene) and three tlr8 (tlr8a1,
tlr8b1 and tlr8b2) genes. Promoter analysis predicted the presence of several transcription factor
binding sites and cytokine regulation of these TLRs. Indeed, tlr7 and tlr8a1 gene expression was influenced by treatment with type I IFN and IFNγ, whereas tlr8a1 and tlr8b1 were most sensitive to treatment with IL-1β [155]. Not all tlr7 and tlr8 genes reacted the same to cytokine treatment and
future studies may want to address putative differences between the duplicated genes. Overall, it is clear that compounds such as RNA-like resiquimod R848 induce type I ifn responses in salmonids, but it remains difficult to unequivocally ascribe this response to sensing by TLR7 and/or TLR8. Certainly, studies on TLR7 and TLR8 in fish species other than cyprinids or salmonids are urgently needed to create a more complete overview.