The Journal of Immunology, 2007, 178: 89.37.
Copyright © 2007 by The American Association of Immunologists, Inc.
TLR5 is not critical for humoral immune responses to purified flagellin or co-administered antigens
Catherine J. Sanders1,
Shizuo Akira2 and
Andrew T. Gewirtz1
1 Dept. of Pathology, Emory University, 615 Michael Street, Atlanta, GA, 30322,
2 Dept. of Host Defense, Osaka University, 3-1 yamada-oka, Suita Osaka, 565-0871, Japan
Abstract
A problem in developing vaccine adjuvants is that their effectiveness may be inherently tied to TLR activation and consequently "cytokine storms". To examine whether this applied to the TLR5 ligand flagellin, a MyD88-dependent adjuvant, we compared immune responses of TLR5KO mice and WT littermates to purified flagellin or flagellin+OVA. Flagellin-treated TLR5KO mice lacked induction of serum IL-12(p40), TNF
, RANTES, IL-17, Eotaxin, or MIP-1
measured over a 24 hour period compared to WT, yet had reduced though significant serum cytokine levels of KC (9%), IL-1
(53%), IL-6 (24%), MCP-1 (22%), and G-CSF (16%) (% KO of WT). TLR5KO mice also lacked detectable activation of splenic DC and had modestly diminished Ab titers to flagellin (26% of WT) or OVA (40% of WT). These data indicate flagellins adjuvanticity does not require TLR5, detectable DC activation, or the relatively high serum cytokine levels found in WT mice. In contrast, MyD88KO mice exhibited complete absence of flagellin-induced cytokines and Abs. Such data suggest the positive antibody response in TLR5KO mice may have resulted from a TLR5-independent pathway of flagellin recognition, whose activation and subsequent immune events promote adaptive immunity. These data support and expand the concept that TLR ligands can promote antibody responses by means not requiring the TLRs thought necessary for their innate immune recognition.