Immunological events induced by intrapulmonary administration of LTK63 or CpG in mice

Summary The synthetic oligodeoxynucleotides containing unmethylated CpG dinucleotides (CpG ODNs) and LTK63, a detoxified mutant of the E. coli the heat labile enterotoxin (LT), are potent mucosal adjuvants. In addition, CpG ODNs and LTK63 provide generic protection, in the absence of co-administered antigen, to respiratory infections. CpG ODNs act through a well-defined molecular pathway, but little is known about immune modulation induced by CpG ODNs in the lung. Similarly, it is clear that LTK63 is immunogenic and functions efficiently as adjuvant, but the mechanism of action in both adjuvanticity and in generic protection is largely unknown. In order to study ex vivo and compare the immunological events induced by CpG ODNs and LTK63, the B-type CpG 1826 or LTK63 were administered intrapulmonary in BALB/c mice. Lungs, sera and spleens were monitored from 3 hours to 14 days after intrapulmonary administration by combining different approaches, comprising multiplex analysis of cytokine protein expression and flow cytometric analysis of lung immune cell populations. In addition, alveolar macrophage (AM) sensitivity to CpG ODNs and the effect of LTK63 on DC recruitment into the lung and on the function of DCs isolated from LTK63 treated and control mice were tested. The cytokine analysis of lung homogenates shows that CpG ODNs induce an early cytokine response and that the earliest cytokines detected in the lung are KC at 1 hr followed by IL-1a and b, IL-12(p40) and IL-6 at 3 hrs and by G-CSF, MCP-1, MIP-1a and b, and RANTES at 6 hrs. Most but not all of these cytokines are found systemically in the serum in a narrow peak at slightly later time points compared to their appearance in the organ, suggesting a spillover from the lung into the blood, and several of those cytokines are released by AMs after in vitro stimulation with CpG ODNs. Flow cytometric analysis of lung immune cells shows that intrapulmonary administration of CpG ODN induces activation of plasmacytoid DCs (pDCs), myeloid DCs (mDCs), CD4 T cells, CD8 T cells and NK cells in a time period of 12 hours to 4 days after treatment, as well as recruitment into the lung of pDCs at 2 days and of mDC starting at 4 days, respectively. LTK63 treatment, in contrast, acts more slowly and induces two phases of activation in the lung: an early phase, which extends from 1 to 2 days, and a second phase, which extends from 6 to 8 days. In the first phase, LTK63 intrapulmonary administration induces up-regulation of IL-1b, G-CSF and KC, which represent granulocyte chemoattractants and growth factors, and consistent with this, accumulation of granulocytes and mDCs in the lung is observed. In the second phase, LTK63 induces in the lung the up-regulation of a complex mixture of cytokines involved in inflammation and cell recruitment, which are produced in part by CD11c+ cells, as shown in in vitro experiments with these cells isolated from the lungs of LTK63 treated mice. Flow cytometric analysis of lung immune cells shows that at 6-8 days after LTK63 treatment, the number of mDCs and pDCs, CD8+ and CD4+ T cells, granulocytes, NK cells and B cells increase, and CD8+ and CD4+ T cell subsets are activated. In vivo migration assays indicate that at least part of the LTK63-induced increase in mDC numbers is due to the recruitment of differentiated DCs from the blood. In order to understand the nature of the increased immune responsiveness in the lung, the T cell stimulatory ability of lung myeloid cells was analyzed. The studies of the mixed population of total lung CD11c+ cells show that cells isolated from LTK63 treated mice are more efficient at stimulating allogeneic T cell responses than those from untreated mice. When we tested the T cell stimulatory ability of cells sorted into CD11c+MHC-IIhigh mDCs and CD11chighMHC-IIint cells comprising alveolar macrophages and immature mDCS, we found no difference on a per cell basis between LTK63 treated and control mice. Since I find a strong accumulation of CD11c+MHC-IIhigh mDCs after LTK63 treatment, I conclude that the enhanced immune responsiveness induced by LTK63 is partly due to increased numbers of CD11c+MHC-IIhigh mDCs with a strong potential to prime T cells. In conclusion, while both CpG ODNs and LTK63 studied here appear to act mainly through APCs such as AMs and DCs, the kinetics of this process greatly differ between the two. The response to CpG ODNs is much faster, probably due to the direct activation of receptor-bearing innate immune cells present in the lung, among which mDCs and pDCs. The earliest event in the response to LTK63, in contrast, is detectable only after 24 hrs, which suggests either that the initial steps are too subtle to be detected here, and that chemokine production and cell influx observed at 1 to 2 days has to be considered the distal result of a sequence of small events preceding those detected here. Alternatively, due to the nature of the interaction between LTK63 and target cells, the phenomena observed at 24 hrs may in fact be the very first to take place, which would suggest a very slow onset of cell activation by LTK63. On the other hand, the response to LTK63 lasts longer, and a second wave of events are observed at 6-8 days. The explanation that is most plausible and compatible with known kinetics of immune responses is that this second wave is driven at least partly by the adaptive immune response to LTK63 and that in fact, LTK63 acts by mimicking infection by a pathogen. At 8 days, we do not detect any response to CpG ODNs anymore, which again would be compatible with the fact that in the absence of a protein Ag component, the immune response is limited to the initial innate phase and not prolonged. This is in line with the studies on generic protection indicating that this protective effect is less durable for CpG ODNs than for LTK63