Skin inflammation mediated by the innate immune system is a crucial step in the sensitization to contact allergens 
. Recent work has shown that contact sensitizers trigger innate immune mechanisms involved in anti-infectious responses 
. These mechanisms are acting in a non-redundant collaborative manner 
. We have shown that CHS induced by organic contact sensitizers such as TNCB and oxazolone is absent in TLR2/TLR4 double deficient mice 
. Since CHS responses develop normally in germ-free mice, we hypothesized that organic sensitizers utilize endogenous skin-derived TLR ligands to generate a pro-inflammatory tissue microenvironment 
. This hypothesis is supported by the fact that TNCB and other organic contact sensitizers up-regulate maturation marker expression by DCs but fail to induce NF-κB dependent cytokine production by DCs in vitro
. Candidates for endogenous TLR2 and TLR4 activators are low MW HA fragments 
. High MW HA prevents TLR2 and TLR4 triggering but inflammation results in the generation of pro-inflammatory low MW HA fragments 
. Combined engagement of TLR4 and the HA receptor CD44 is needed for a full inflammatory response to HA fragments in vitro
A putative role for HA in CHS has been described by us recently 
. Blocking HA function with a peptide inhibitor prior to sensitization significantly reduced the CHS response to TNCB in germ-free mice. Therefore, we assumed that also in CHS high MW HA must be degraded to low MW fragments that provide necessary endogenous activators of innate inflammatory signaling in the skin.
We now show the degradation of HA in the epidermis of mice following application of contact sensitizers. This may trigger the release of soluble low MW HA and promote inflammation as described 
. Although to our knowledge no direct interaction of HA with TLR2 or TLR4 has been demonstrated so far, our data provides strong in vivo
evidence that the degradation of HA to low molecular weight fragments is essential for the induction of a pro-inflammatory tissue micromilieu. In combination with the data that TLR2/4 expression is crucial for the induction of CHS our data therefore suggests that the inflammation caused by HA degradation involves TLR signaling. So far, we have unfortunately been unable to show an in vitro
activation of DC using commercial low molecular weight HA fragments ranging from 2 to 12mers obtained from different suppliers, none of these fragments showed DC activating capacities (data not shown). This may be due to the fact that the exact HA fragment size required for TLR2/4 activation has not been clearly defined using synthetic material. Moreover, the activating structure in vivo
may be different from the synthetic fragments, for example it may contain HA binding proteins as found in the ECM. Previous publications showing in vitro
activation of DC and a role for HA in lung injury have all used their own HA preparations which make a direct comparison of the data difficult and leave open the possibility that the HA fragments from biologic material might contain either other TLR2/4 ligands or ligands for other TLRs. Biochemical data on the direct interaction of fragmented HA with TLR2/4 are urgently needed.
HA fragments can be generated either by the activity of HA degrading enzymes, the hyaluronidases, or by oxidative de-polymerization induced by ROS 
. As HA degradation also occurs in human skin when contact sensitizers are applied ex vivo,
our data indicate for a species spanning mechanism.
Inhibition of ROS induced HA degradation was linked to a reduction of inflammation in both bleomycin- and asbestos-induced models of pulmonary fibrosis 
as well as in a reduction of inflammatory gene expression in alveolar macrophages and epithelial cells (35). Remarkably, contact sensitizers induce ROS production both in human keratinocyte cell lines 
as well as in DCs 
We demonstrate here that contact sensitizers induce ROS dependent degradation of high molecular weight HA in the skin. Our data did not show clear inhibitory effects of APDC treatment or gp91phox deficiency. This suggests a role for ROS from different cellular sources most likely including NADPH-oxidase dependent and mitochondrial ROS production. It remains to be determined whether these different ROS sources are redundant, additive or synergistic. In addition, we detect upregulation of hyaluronidase activity by contact sensitizers in the skin that promotes HA degradation. Functional inhibition of hyaluronidases by the hyaluronidase inhibitor AA abrogates not only trauma induced IL-6 production in vitro,
but, more importantly, prevents sensitization for CHS. This inhibition is reverted when exogenous active hyaluronidase is co-administered with AA. Interestingly, inhibition of p38 MAPK activation also prevents CHS responses. This effect is at least in part dependent on the activity of hyaluronidases as co-administration of exogenous hyaluronidase reverts the inhibitory effect of the p38 MAPK inhibitor. These results are in line with a recent study showing that activation of p38 MAPK in lung inflammation results in enhanced hyaluronidase activity which in turn leads to the generation of low MW HA fragments and exaggerated inflammation 
So far, the functional role of HA degradation in ACD is not fully understood. In the present study, we underscore the pro-inflammatory role of HA breakdown in CHS and show that contact sensitizers modulate HA metabolism. Our data support the concept that contact sensitizer induced DAMPs serve as endogenous danger signals that are perceived by innate immune receptors 
. Thus, contact sensitizers induce HA breakdown which may result in TLR2-, TLR4- and CD44-dependent DC activation. In CHS, this HA mediated signal is delivered in the tissue microenvironment of the skin and is required for full DC activation in addition to the TLR independent induction of co-stimulatory molecules such as CD86 
. Our data are in line with recent reports showing that congenital over-expression of hyaluronic acid synthetase 2 (HAS2) in Shar Pei dogs results in reoccurring breakdown of HA into low MW fragments, leading to the inflammatory hereditary periodic fever syndrome 
HA breakdown in CHS seems to be initially ROS mediated given the rapid induction of ROS by contact sensitizers. The central importance for ROS in chemical induced skin inflammation has been highlighted by the prevention of CHS involving a block of Langerhans cell migration upon keratinocyte directed over-expression of extracellular superoxide dismutase 
. This may be due to the prevention of the oxidative and enzymatic HA degradation in the absence of sufficient amounts of ROS since ROS also regulate p38 MAPK dependent up-regulation of hyaluronidase as recently described for lung inflammation 
Thus, contact sensitizers induce endogenous danger signaling by triggering ROS and hyaluronidase mediated HA degradation. This process is crucial for CHS since sensitization is completely prevented by pre-treatment of the skin with antioxidants or the hyaluronidase inhibitor AA. In both cases, the inhibitory effects can be overcome by active hyaluronidase. However, at least regarding IL-6 production in ear sheets, the trauma induced IL-6 production occurring due to the mechanical separation of the ear sheets seems to result in a maximal cytokine production that can neither be significantly enhanced by addition of TNBS nor by hyaluronidase. Most interestingly, antioxidant application also prevents elicitation of CHS in sensitized animals. The role of HA induced signaling in the challenge phase of CHS remains to be determined. In vitro signaling studies with cells from contact allergen sensitized skin are, however, hampered by background problems due to the cell isolation procedure. Moreover, the deletion of the tissue context may significantly change cell behavior. Therefore, in vivo approaches using signaling inhibitors as initiated by us for p38 MAPK may provide more relevant results that could be translated into the development of novel treatment strategies.
Our findings add to an emerging, more general scheme highlighting an important functional role for ECM components as endogenous regulators of inflammation 
. Enhanced (or exaggerated) ECM degradation that disturbs the homeostasis of ECM turnover signals danger to the innate immune system. HA and biglycan are implicated in the activation of TLR2 and TLR4 signaling as well as in inflammasome activation 
. As shown here, blocking HA breakdown can prevent sensitization for CHS. This finding encourages the search for inhibitors of ECM degradation or for antagonists of the pro-inflammatory function of ECM components, which can be used in the prevention and therapy of inflammatory skin diseases such as ACD.