Plasmacytoid dendritic cells are well known for their robust type I interferon response to viral infections, and for their ability to induce tolerance via Treg induction. In this study we sought to address a poorly understood facet of pDC biology; their role in immune responses to bacterial infections. Here we found that depletion of pDCs during a Chlamydia pneumoniae infection in mice resulted in delayed lung inflammation and bacterial clearance, with a significant reduction in early cytokine production. This resulted in an increase in a late chronic inflammation characterized by cellular infiltrates, iBALT like structures containing large numbers of B and T-cells, and increased IL-12.
One caveat to the antibody depletion model of pDCs (BST2 based) is the possible depletion of other activated immune cells 
. However we have addressed the specificity issue by confirming our findings in the DT depletion model using the recently described BDCA2-DTRtg mice 
. While the BDCA2-DTRtg mice had similar reductions in cytokine production and BALF cell counts compared with the Ab depleted mice, the changes in cell numbers were not as impressive. However, the BDCA2-DTRtg mice also appeared to have a higher baseline inflammation as assessed by BALF cell counts, thus making direct comparisons between the two models more difficult.
Very little is known about the role of pDCs during bacterial infections and their function in innate immune responses are just being investigated. In one study, pDCs were able to stimulate CD8 T-cell proliferation and IFN-γ production during a Listeria monocytogenes
. pDCs were also required during a CpG induced protective immune response to Listeria
. However, using a DTR ablation model, Takagi et al
found that pDC depletion increased survival during Listeria
infection by reducing overall inflammation 
. Additionally, in a short study published by Ang et al
, pDCs were induced during Legionella pneumophila
infection and pDC depletion led to a delay in clearance of the bacteria, but its effect on inflammation was not described 
We had previously shown that the numbers of pDCs increased in the lungs of CP infected mice 
. Similar to Ang et al
, we found that pDC depletion resulted in a delayed bacterial clearance 
. However, we also saw a dramatic decrease in inflammation early during lung infection with CP. While Takagi et al
also found a decrease in inflammation during Listeria infection and pDC depletion 
, this was actually beneficial to the host. However, in our study, depletion of pDCs lead to delayed bacterial clearance and a prolonged chronic inflammation after CP infection. Thus depending on the type of bacterial infection, pDCs may play different roles and their depletion may result in beneficial or detrimental effects to the host.
Whether pDCs in the lung respond to CP infection directly, or by some secondary signal is currently unknown. CP infection is detected by both TLR2 and TLR4 
, however, pDCs generally lack surface Toll-like receptors 
. It is currently unknown if pDCs can sense CP via NOD/Rip2 signaling, nor has there been any studies indicating TLR9 detection of CP infection. We have found that bone marrow derived pDCs can make IL-12 in response to CP infection but the mechanism is not known (data not shown).
While pDCs can make IL-12 themselves, the significant reduction in IL-12p40 levels seen in pDC-depleted animals early during CP infection is unlikely due to any IL-12 the pDCs produced themselves as there are relatively low numbers of pDCs present in the lungs. Thus it is more likely that pDCs provide some other signal that induces macrophages and cDCs to produce larges amounts of IL-12. pDCs are known for their ability to make type I interferon, however, one study found that mice that lacked the type I IFN receptor, if anything, cleared CP infection better than WT mice 
. Thus it is unlikely that our observations during pDC depletion are due to any involvement of type I IFN. pDCs are also known to influence Tregs via the production of indoleamine 2–3 dioxygenase (IDO) 
. However, mice fed 1-methyltryptophan (an IDO inhibitor) during CP infection did not differ from control mice in their immune responses 
. Another possible mechanism could be that pDCs are required for cross-presentation to naive CD8 T cells 
. Indeed we found a reduction in CD8 T cells early during CP infection in the lungs of pDC-depleted animals. However, there were increased CD8 T cells late during infection. Finally, pDCs have also been implicated in NK cell activation 
, a crucial source of IFN-γ early during CP infection. However, we did not find any significant changes in these cells during pDC depletion (data not shown).
Björck et al
recently found that pDCs can come in two generic phenotypes, one that is proinflammatory, and another that is more suppressive 
. In our study we found that while inflammation and cytokine production were reduced in the lungs of pDC-depleted CP infected mice, an opposite result was observed in the draining lymph node. Restimulation of DLN cells (5 days after infection) in pDC-depleted animals resulted in increased IFN-γ production, an apparent paradox given the reduced inflammation seen in the lungs. However, similar to what Björck et al
found, lung pDCs and DLN pDCs had different levels of proinflammatory and suppressive surface markers during CP infection, coinciding with their respective phenotypes. This highlights the complexity of studying the role of specific cell types using a depletion model. Perhaps the late increase in CD4 and CD8 T-cells in the lung that we observed in pDC-depleted animals is due to enhanced T-cell activity in the DLN as a result of the depletion of suppressive pDCs in the DLN. These questions would be difficult to address, however, as the proinflammatory version of pDCs can switch to the suppressive phenotype 
Perturbation of early innate immune responses to CP lung infection tends to results in an initial delay in inflammation and bacterial clearing, followed by an increase in chronic inflammation later on via secondary pathways that is poorly understood 
. pDC depletion also resulted in an early delay in cellular recruitment and defective cytokine production with hindered bacterial clearance. This lead to a profound delayed and chronic inflammation, characterized by increased inflammatory cells (especially CD8 T-cells), high levels of IL-12, iBALT (B-cell and T-cell positive) like structures in the lungs of pDC-depleted CP infected animals. iBALTs are thought to provide an enhanced local immunity against the offending pathogen 
. However, the origin and function of these CP infection-induced structures are currently not known and are under intense investigations. While we do not know by what mechanism the late inflammation is generated, it is likely that it is either due to the significant delay in bacterial clearance and secondary responses, or a dysregulation in immune responses due to prolonged pDC depletion. Considering the associations between CP infections and chronic airway diseases such as asthma and COPD, understanding the development and effects of this prolonged inflammation is a high priority. pDCs clearly play a role for proper host immune responses to CP infection (that does not require type I IFN), suggesting a much greater influence on innate immune responses to bacterial infections, however, the mechanisms for these interactions still needs to be established.