Introduction

Human immune responses are mediated by innate and adaptive immunity. Innate immunity is the first line of defense against pathogens, whereas adaptive immunity raises pathogen-specific responses. Cells that confer innate immunity to humans can detect various pathogens using pattern recognition receptors (PRRs).1 PRRs are a class of receptors that recognize specific molecular patterns in pathogens, such as pathogen-associated molecular patterns (PAMPs) and trigger their respective immune responses.2 Most characterized PAMPs are nucleic acids of pathogens, such as single-stranded (ss) and double-stranded (ds) DNA and RNA. These PAMPs activate PRRs to produce type-1 interferons and proinflammatory cytokines to combat pathogens. In humans, only a few well-known PRRs are present which includes toll-like receptors (TLRs),3 nucleotide-binding oligomerization domain-like receptors4 and retinoic acid-inducible gene I-like receptors.5 It is important that PRRs precisely recognize nucleic acids from pathogens and avoid self-recognition of host nucleic acids. This precise recognition is achieved by regulation, such as localization and processing of pathogen nucleic acids, post-translational processing of PRRs, and specific interactions of PRRs with nucleic acids.2 These PRR regulations are crucial for understanding the mechanisms of the immune response, autoimmune diseases, as well as the development of vaccines and other therapeutics.1

One of the well-known PRR members is the TLRs. TLRs are transmembrane proteins that are evolutionarily conserved between insects and vertebrates.3 These receptors are essential for the innate immune system as they recognize PAMPs, such as lipopolysaccharides, peptidoglycans, flagellin, and unmethylated cytosine-phosphate-guanine (CpG) DNA.6 The activation of TLRs leads to a signaling pathway involving the transcription factors NF-kB and AP-1. These transcription factors work synergistically to initiate cytokines, chemokines and matrix-degrading enzyme for bacterial and viral infections. There are 10 TLRs recorded in humans,7 13 in mice8 and at least 6 (TLRs 2–4 and 7–(9) in horses.9,10,11 Among TLRs, human TLR3, TLR7/TLR8, TLR9 and mouse TLR13 recognize nucleic acids and are localized in the endosomal compartment. TLR3, TLR7/TLR8 and TLR9 sense dsRNA, ssRNA and CpG DNA respectively12,13 while, mouse TLR13 binds to 23 ribosomal RNA.14,15 TLR7 and TLR8 can recognize ssRNAs and TLR9 is specific to single-stranded DNA (ssDNA) with a CpG dideoxynucleotide motif. Since TLR7, TLR8, and TLR9 recognize single-stranded nucleic acids, they are grouped under the TLR7 subfamily. These TLRs have the same number of leucine-rich repeats (LRRs).

As mentioned earlier, TLR9 in the TLR7 subfamily recognizes unmethylated CpG motifs in ssDNA of bacteria and viruses and is initially localized in the endoplasmic reticulum (ER) before exiting to the endosomal compartments.16 The ectodomain of TLR9 is oriented toward the lumen of endosomes, while its C-terminal domain extends into the cytoplasm. Multiple mechanisms maintain the proper localization of TLRs, including motifs in the cytoplasmic tail that retain TLR9 in the ER and TLR9 phosphorylation.17 TLR9 also contains LRRs i.e., LRR1–2618 and between LRR14 and LRR15 lies a long loop region of 40–50 amino acids (aa) called the Z loop. Several studies have shown that the Z-loop cleavage is essential for TLR7 subfamily activation (Figure 1A).18 Structurally, TLR9 comprises an N-terminal LRR (26–61 aa), LRR1–14 (62–440 aa), Z loop (441–469), LRR15–26 (470–772 aa), C-terminal LRR (773–818 aa), a transmembrane segment (819–839 aa), and a cytoplasmic toll/interleukin-1 receptor/resistance protein (TIR) domain (840–1,032 aa).19 Cleaved TLR9 contains 471–1,032 aa, spanning the LRRs (15–26) to the C-terminal end.20 TLR7 and TLR8 simultaneously bind two ligands: a ssRNA and a mononucleoside for dimerization.21,22,23 On the other hand, The TLR9 dimer harbors two ligand binding sites: one for the CpG ssDNA and the other for the 5′-xCx DNA. TLR9 dimerization is enhanced by 5′-xCx DNA in the presence of CpG DNA, consistent with synergistic ligand engagement.24,25 Crystal structures of Equus caballus TLR9 show that two receptor ectodomains bind two DNA ligands to form the active complex, and this architecture is highly relevant to human TLR9 given the >80% sequence conservation in the ligand-binding and dimerization regions ligands.24,25 Mouse TLR9 structures likewise reveal a dimer engaging two DNA molecules, indicating that this stoichiometry is conserved across mammals.21 Functional studies in human TLR9 further demonstrate that TLR9 requires two CpG motifs, consistent with a bivalent ligand-receptor arrangement.26,27,28,29 Pohar et al. also showed that the short ssDNA containing a 5′-TCG motif augments CpG-containing DNA-induced TLR9 structural rearrangement.28

Figure 1: TLR9 complexStructure of CpGDNA- and IM-bound TLR9 complex (A). Evolutionarily conserved regions of TLR9 (B).

Multiple studies have been conducted on mouse, bovine, and horse TLR9; however, little is known about dynamics and key interactions between human TLR9, CpG ssDNA, and IM complex.10,21,30 The crystal structure of human TLR9 has not yet been elucidated but the advancement in computational tools now allows us to model a protein of interest using homologous structures. In this study, we modeled human TLR9 based on the crystal structure of Equus caballus TLR9 (Figure 1A) and simulated TLR9 in complex with CpG ssDNA and a 5′-xCx DNA immunomodulator (IM) to understand the dynamic conformational changes in TLR9 and the mechanism of ssDNA ligand engagement. We report various intra-, inter-monomer, and also TLR9-CpG-IM interactions that might help design ligands for therapeutic purposes.