Important mutations do not always declare themselves without challenge. This is particularly true of mutations that affect immune function. Without an infection, immunocompromised mice may seem entirely normal. For example, the C3H/HeJ mouse is indistinguishable from other C3H substrains unless it is challenged with LPS or an authentic Gram-negative infection (6). It may be that many of our genes are conditionally important in the same way. In the Beutler lab, genetic screens are presently being applied to study four important topics in immunobiology.
Signaling pathways utilized by the TLRs and other innate immune sensors are kept under surveillance in screens designed to detect mutations that impair the detection of microbes.
In the TLR signaling screen, signaling from seven TLRs is monitored by measuring tumor necrosis factor (TNF)-α production by peritoneal macrophages from ENU-mutagenized mice ex vivo. This screen has led to the decipherment of pathways for microbe sensing, identifying proteins that could not be "guessed" to participate in signaling (8-10). In addition, the study of several mutants identified in the screen has revealed subtleties in the nature of signaling from several TLRs (9, 11, 12). For example, the pococurante mutation of MyD88 demonstrated that signaling from TLR2 is inherently different from signaling through the other TLRs, requiring only one of two known sites of receptor-adapter interaction(Figure 3)(11). The Double-stranded DNA Macrophage Screen, to identify components involved in sensing cytoplasmic double-stranded DNA (dsDNA), and the NLRP3 Inflammasome Screen, to identify components involved in sensing “danger signals,” are also being carried out on macrophages ex vivo. An in vivo screen for response to injected CpG oligodeoxynucleotides has recently been initiated.
Infecting mice with authentic pathogens using small inocula that are normally eliminated or contained by mice, can detect mutations that impair host defense.
Screens for susceptibility to mouse cytomegalovirus infection (MCMV Susceptibility and Resistance Screen), and for clearance of lymphocytic choriomeningitis virus (LCMV Clearance Screen) in vivo are currently underway. These screens rely on the highly reproducible behavior of mice challenged by infection, which assures that phenovariants may be easily discerned. Some of the identified mutations have also come as great surprises (13). For example, mayday mice die between 24 and 72 hours after infection with 5 x 104 PFU of MCMV, and were found to carry a mutation in the gene encoding an inwardly-rectifying potassium (K+) channel subunit, Kir6.1 (14). Screens for control of MCMV, adenovirus, influenza, and Rift Valley Fever Virus are performed in macrophages ex vivo (Ex Vivo Macrophage Screen for Control of Viral Infection).
ENU mutations can also render mice highly resistant to infection by specific pathogens, or result in autoimmune and inflammatory disease.
The DSS-induced Colitis Screen is designed to discover mutations that result in susceptibility to chemically-induced colitis, which is thought to arise from excessive and sustained inflammatory host immune responses against commensal intestinal microbes. The screen monitors weight loss as an indication of colitis (Figure 2) allowing sensitizing mutations to be easily retrieved. Looking for changes in DSS sensitivity may also reveal mutations that inappropriately activate immune responses to normal intestinal flora. Because of their potential to activate both innate and adaptive immune systems, mutations identified in each of these screens may also reveal molecules that contribute to autoimmune disease. (Figure 3)
The nature of the innate:adaptive immune connection is being probed.
Although the innate immune response clearly contributes to the development of an adaptive immune response, the mechanism by which this occurs remains unclear. Together with our colleagues in the Nemazee lab at The Scripps Research Institute, we have shown that TLR signaling is not required for both effective antibody production following immunization with any classical adjuvant, including Freund’s adjuvant (15), or for strong cytotoxic T-lymphocyte (CTL) responses (16). Using a viral vector (recombinant Semliki Forest Virus; rSFV) to produce a model antigen, we have probed T-dependent B cell responses. In addition, we have used NP-Ficoll to probe T-independent B cell responses and flow cytometry to look for defects of immune cell development. We have found a total of 19 mutations that impair the antibody response. Interestingly, all of them affect cells of the adaptive immune system rather than the innate immune system (17).