The Danger Model, and how tissues control Immunity

National Institute Of Allergy And Infectious Diseases

Investigators

Linked publications, trials & patents

Abstract

We used six different experimental set-ups[unreadable] [unreadable] 1) BIODEFENSE AND NEONATAL IMMUNITY Although there is a concerted national effort to find vaccines for potential bio-terrorism agents, little is being done to protect the nation's infants. Most adult vaccines do not work for infants less than 6mo old, and many do not work for babies < 1yr. Because we cannot leave our children behind, we have been analyzing the immune responses of newborns.[unreadable] A) The inhibitory effect of maternal antibodies: In 1996 we showed that newborn mice were immunologically competent (contrary to common opinion), as long as they were properly immunized (Science 271:1723-1726). We have now been looking at a second myth, namely that the presence of maternal antibody inhibits newborn responses. in the last two years, we found that neonatal mice respond perfectly well to the model antigen Ovalbumin (OVA), even if they have received passive anti-OVA antibody from their mothers. The immunity in babies from immunized mothers lasts for at least 18 months, and does not fade more rapidly than that of babies from control mothers. Thus, in mice, maternal antibodies are not inhibitory. [unreadable] B) anthrax: To see if our results with the model antigen, OVA, hold for real infections, we studyied the response to Anthrax. In preliminary studies to find the right vaccine dose, we found (astonishingly!) that the current test for adult immunity (neutralizing antibody) is misleading. Mice immunized with Anthrax protective antigen (PA) in adjuvant make high titers of antibody (as seen on ELISA tests) but no neutralizing antibody. Yet they are completely protected from challenge with live Anthrax. This result may radically change how to vaccinate individuals and will also change how we analyze the responses of newborn mice. Further, we found that neonates immunized with Anthrax PA in adjuvant are protected from challenge, though some of them make no detectable antibody at all, suggesting that non-antibody protective mechanisms exist.[unreadable] C) Neonatal responses to Measles: having found that neonatal mice respond perfectly well to both OVA and Anthrax, we turned to the measles vaccine to ask why neonatal humans do not respond well. We have been studying the vaccine variant of the Edmonston strain of Measles to see what properties make it a poor vaccine for children. We found that the CD46 isotypes expressed by cord blood are not the same as those expressed by adults, and that the neonatal isoforms are not conducive to infection by the vaccine Measles strain.[unreadable] [unreadable] 2) CD4 T CELLS CLEAR TUMORS We found that CD4 T cells can be better at clearing tumors than CD8 cells. Against six out of six tumors, from five different tissues, CD4 effectors were more potent than CD8s. Searching for the mechanism, we found that the CD4 T cells partner with NK cells. We are now studying this partnership to determine which cell does what. Thus far, it seems that IFN-gamma production by the CD4 T cells is involved, and that tumor infiltrating macrophages can be educated by CD4 T cells to kill tumor cells. (Perez-Diez, et al (2007) CD4 Cells Can Be More Efficient at Tumor Rejection Than CD8 Cells Blood 109:5346[unreadable] [unreadable] 4) DENDRITIC CELLS & IL-12p70 It is currently thought that dendritic cells become exhausted within 24 hours of stimulation by LPS. We have found that this is incorrect. The dendritic cells, while resistant to restimulation by LPS, can produce IL-12p70 when re-stimulated by activated T cells. We found that TH1 cells and TH0 cells can stimulate this production of IL-12p70 while TH2 cells cannot. The 'exhausted' dendritic cells make a host of cytokines and chemokines not made by freshly stimulated dendritic cells. Thus dendritic cells are not in control of immune effector class, but instead relay signals from the cells with which they are in contact. [unreadable] [unreadable] 5) IMMUNITY & HEALING: The Danger model led us to the view that tissues influence the effector class of immunity in order to prevent immune-mediated local damage. To determine whether wounded tissues are affected by the immune system, we punched small holes in the ears of mice and measured the healing rate. We found that young mice heal slowly while older mice, surprisingly, quickly regenerate both the epithelium and cartilage so well that there is eventually no visible scar. Histology suggested that the amount of inflammation is different between the two ages of mice. Thus we tested NSAIDs and found that they greatly inhibit the regeneration, while croton oil enhances it. We tested many mutant strains of mice and found that there is essentially no difference in the regenerative capacity of middle aged mice of most strains. [unreadable] [unreadable] 6) TOLERANCE TO LATE-APPEARING ANTIGENS: One of the problems with the self-non-self model is that it does not account for tolerance to antigens that appear late in life, such as the milk proteins of lactating mothers (eg. alpha lactalbumin, beta lactoglobulin, casein etc.). To see if tolerance to these proteins is established in the thymus, or perhaps by fetal exposure during pregnancy or neonatal exposure during lactation, we have been using a model system in which we can obtain adult animals that carry a particular milk-protein gene but which have not previously encountered the protein, either across the placenta, or by drinking the milk as babies. [unreadable] [unreadable] HUMAN ALPHA LACTABLUMIN KNOCK-IN MICE: We obtained mice that had their own alpha lactalbumin (mALAC) genes replaced with the human allele (hALAC), and bred them for 22 generations to B10.BR. To make experimental animals, we bred normal B10.BR females to B10.BR hALAC males. The F1 progeny carry both human and mouse alleles of ALAC but have only been exposed to mALAC from their mothers. When immunized to hALAC after weaning but before puberty, these mice (both males and females) made good T cell and antibody responses. Thus they were not rendered tolerant by thymic expression. When bred, the immunized females showed no autoimmune reactions to their mammary tissue. Neither did the levels of antibodies to hALAC drop or the subclasses of antibodies change. Using pep-scan, we found that the carriers, which make good responses, are nveretheless somewhat tolerant, as they do not respond to a dominant peptide that stimulates non-carrier mice quite well. We made T-cell hybridomas from immunized normal female mice, to make TCR Tg mice from these to follow the fate of immunized self-reactive T cells as the mice lactate through several pregnancies. [unreadable] [unreadable] 7) THE EFFECT OF COMMENSAL BACTERIA ON THE HEALTH OF INTESTINAL TISSUE: we are studying, by microarray analysis of lazer microdissected sections of small intestine, the response of various tissues (eg lamina propria, epithelium, crypts) to the presence or absence of commensal bacteria. Preliminary data suggest that many more cells produce anti-microbial peptides than previously thought.