MD-1 and MD-2 are secretory glycoproteins that exist on the cell surface in complexes with transmembrane proteins. MD-1 is anchored by radioprotective 105 (RP105), and MD-2 is associated with TLR4. In vivo studies revealed that MD-1 and MD-2 have roles in responses to LPS. Although the direct binding function of MD-2 to LPS has been observed, the physiological function of MD-1 remains unknown. In this study, we compared the LPS-binding functions of MD-1 and MD-2. LPS binding to cell surface complexes was detected for cells transfected with TLR4/MD-2. In contrast, binding was not observed for RP105/MD-1-transfected cells. When rMD-2 protein was expressed in Escherichia coli, it was purified in complexes containing LPS. In contrast, preparations of MD-1 did not contain LPS. When rMD-2 protein was prepared in a mutant strain lacking the lpxM gene, LPS binding disappeared. Therefore, the secondary myristoyl chain attached to the (R)-3-hydroxymyristoyl chain added by LpxM is required for LPS recognition by MD-2, under these conditions. An amphipathic cluster composed of basic and hydrophobic residues in MD-2 has been suggested to be the LPS-binding site. We specifically focused on two Phe residues (119 and 121), which can associate with fatty acids. A mutation at Phe191 or Phe121 strongly reduced binding activity, and a double mutation at these residues prevented any binding from occurring. The Phe residues are present in MD-2 and absent in MD-1. Therefore, the LPS recognition mechanism by RP105/MD-1 is distinct from that of TLR4/MD-2.

Ligands for toll-like receptors (TLR) are known to induce a variety of immune responses. Selective induction of desirable responses would be important for the treatment of individual diseases with various pathogenesis. For this purpose, we established six MAbs against the TLR4/MD-2 complex (UT MAbs) from TLR4(-/-) mice or MD-2(-/-) mice. Three MAbs (UT12, 18, and 22) induced NF-kappaB activation and production of pro-inflammatory cytokines, but the other three (UT15, 41, and 49) did not induce such cell responses. Unlike lipopolysaccharide (LPS), agonistic UT MAbs did not require serum components for the functions. UT41 and UT49 recognized TLR4 in the absence of MD-2. On the other hand, the other four MAbs reacted to the TLR4/MD-2 complex, but not to solo TLR4. Agonistic UT MAbs shared the epitopes, but non-agonistic UT15 reacted to distinct epitope on the complex. UT MAbs appear to be useful analyzing the molecular mechanism of TLR signaling and will contribute to the development of novel immunotherapies.

No abstract

No abstract

No abstract