Over the last few years evidence has accumulated indicating that innate immune responses of newborns and adults differ significantly. Several mechanisms may contribute to these observations. It is therefore difficult to identify the developmental processes that will lead to the acquisition of adult-like responses. Herein, we focused on specific aspects of neonatal TLR responses that have previously been analyzed using cord blood. As summarized in , we analyzed the ontogeny of these parameters over the first year of life. A consistent finding is the decreased capacity of cord blood-derived cells to produce bioactive IL-12p70 by monocytes and mDCs. In response to LPS, we detected low but reproducible IL-12p70 production in adult samples. By the age of 6 months, adult levels of IL-12p70 were observed in some LPS-stimulated samples. Our data is in line with a recent report that indicates that LPS+IFNγ-induced IL-12p70 production in whole blood was significantly increased between birth and 1 month of age but still low compared to adult samples 
. A previous report indicated that decreased capacity to produce IL-12p70 was observed throughout childhood (samples from 5- and 12-yr-old children were analyzed) 
. There are major differences in the design of the experiments that could account for this apparent discrepancy. In their work, Upham et al
analyzed IL-12p70 production in isolated PBMC, cultured in FCS-containing medium and stimulated by LPS and IFNγ combination. Here, we analyzed IL-12p70 production in LPS-stimulated whole blood. Multiple parameters that differ between these experiments could affect the capacity of APCs to produce IL-12p70. For example, circulating plasma factors and the presence of red blood cells affect cytokine production 
. The cellular source of IL-12 could also differ. Indeed, when monocytes are primed with IFN-γ, they gain the ability to produce IL-12p70 through upregulation of IRF1 and IRF8 expression 
. Taken together, these data indicate that the capacity of circulating APCs to produce IL-12p70 in response to LPS gradually increases within the first months of life. However, this might not reflect the capacity of cells to produce IL-12 under more intense stimulation conditions. Reduced production of IFN-γ in response to LPS was observed throughout the first year of life. This is consistent with the fact that it reflects both the capacity of TLR4-expressing cells to produce IL-12 and of lymphocytes and NK cells to produce IFN-γ. Indeed, upon direct stimulation of lymphocytes by phytohemagglutinin, production of IFN-γ was found to be low in children until the age of 18 months 
Summary of the main results.
High circulating adenosine levels contribute to the low capacity of cord blood cells to produce TNF-α in response to LPS 
. We observed that by the age of 6 months, TNFα production reached adult levels, suggesting that alteration in plasma factors last for more than 3 months after birth. Differences in plasma factors also contribute to low IP-10 production in LPS-stimulated cord blood. However, we showed that cell intrinsic factors are also important. Indeed, we previously showed that incomplete IRF3 activation in cord blood cells led to decreased IFNβ production and induction of IFN-dependent genes such as IP-10 
. IP-10 production reached adult levels by the age of 9 months. This result suggests that the differences in signalling pathways are gradually overcome before that age.
We previously reported that neonatal pDCs produce less type I IFNs 
. As a consequence, IFN-inducible genes, such as IP-10 and MIG levels are strongly reduced upon TLR9 stimulation of cord blood. With age, we observe a gradual increase in production of these 2 chemokines. However, in 1-year old infants, levels were still significantly lower than in adults. This result strongly suggests that pDC reach adult-like function later in life than monocytes or mDCs. We observed a similar trend for phenotypic markers ().
As previously reported 
, we observed high production of IL-6, a pleiotropic cytokine, in cord blood samples as compared to adult samples. LPS-induced IL-6 production was comparable to adult levels by the age of 3 months. In contrast, we detected high IL-8 and IL-10 production in some samples from 3- to 12-month old infants. We observed a dramatic increase in IL-6, IL-8, IL-10 and IL-1β production in CpG-stimulated cord blood but not samples from older infants or adults. Interestingly, production of these cytokines by purified cord blood pDCs was not found to be increased 
. This result suggests that an alternative cellular source of these cytokines could be more responsive to TLR9 stimulation (in a direct or indirect fashion) in cord blood. Cytokine production was found to be affected by the mode of delivery 
. However, no differences in CpG-induced cytokine levels were observed between children that were delivered naturally or through C-section. This transient overexpression of these inflammatory cytokines should be kept in mind in the context of vaccination of young infants with TLR ligand-containing adjuvants, such as monophosphoryl lipid A (MPL).
In conclusion, our study indicates that most of the parameters of TLR responses we analyzed progressively reach adult-like levels within the first year of life. Clearly, intra-uterine environment conditions the function of APCs. It remains unknown whether “maturation” of these responses reflects the normal turn-over of APCs and the disappearance of these immunomodulators or if it is an active process that requires education of immune cells by environmental exposure to microbial compounds. It would therefore be helpful to assess the impact of specific factors (vaccination, intercurrent infections, atopic background etc.) and settings (resource-rich vs. developing countries) on the ontogeny of TLR responses in early life.