Intrauterine infection is a common cause of preterm delivery. Chorioamniotitis or inflammation in the fetal membranes is associated with over 60% of preterm deliveries 
. The most common organisms isolated from the amniotic fluid of women with chorioamnionitis are the Ureaplasma
. These organisms are commonly found within the uterus in association with spontaneous preterm labour and with preterm PROM. In addition, the intrauterine presence of these organisms has been linked with an increased production of a wide variety of cytokines, matrix metalloproteinases and postaglandins, all believed to be in the causal pathway for PROM, thus activation of the innate immune system has been suggested 
are unusual bacteria in that they have a plasma membrane but lack a peptidoglycan cell wall, use urea as the sole source of energy, and are dependent on the host for other metabolic functions 
. Since they lack a conventional bacterial cell wall, the question that remains is how they trigger the innate immune system and which PRRs they engage. It is imperative to elucidate how Ureaplasma
species can induce an inflammatory response, if we are to understand how these bacteria cause PROM.
Although much attention has been paid to TLR expression and function in the female genital track 
, little is known about TLRs in the human amniotic epithelial cells and how they recognise Ureaplasma
species. In this study, we set out to identify which TLRs are involved in triggering pro-inflammatory responses in response to Ureaplasma
species in human amniotic epithelial cells. Initially we investigated whether Ureaplasma
can trigger the secretion of pro-inflammatory cytokines in human amniotic epithelial cells. Our experiments suggest that Ureaplasma SV2, SV3
are capable of inducing the secretion of pro-inflammatory cytokines, which include TNF-
IL-6, IL-8 and IL-1β within the first two hours of their interaction with the host. No significant difference was observed among the different Ureaplasma
serovars. Thus we proceeded to elucidate the mechanism by which Ureaplasma
can induce a cytokine response in human amniotic epithelial cells. Since TLRs are the key receptors for sensing bacteria, we investigated whether TLRs were involved in the cytokine production in response to Ureaplasma
. In order to examine which if any, of the TLR molecules are involved, we utilised HEK293 cells transfected with different TLRs. It was shown that Ureaplasma
species were able to activate only cells transfected with either TLR2, TLR2/6 or TLR9 and to a lesser extent TLR4. This is partly in agreement with Shimizu et al. 
where they suggested that TLR1, TLR2 and TLR6 are involved in sensing Ureaplasma parvum
. In our study, we did not find any TLR1 involvement, whereas we found TLR9 involvement. Differences in the two studies might lie in the fact that the Shimizu et al study was performed with bacteria cultured in media supplemented with yeast extract, which itself constitutes a PAMP, thus results obtained might have been affected by the presence of yeast components.
Furthermore, similarly with the study by Shimizu et al. 
we proceeded to investigate whether lipoproteins were triggering the response observed. It was shown that MBA was triggering responses via TLR2/6, thus suggesting that this heterodimer might be engaged on the cell surface prior to Ureaplasma
The ability to induce an inflammatory response via TLR9 seems to suggest that Ureaplasma species must internalise and replicate within the host's cells. Thus suggesting that the innate recognition of Ureaplasma in the early stages of attachment to the host cells is via the recognition of lipoproteins by TLR2/6, but once the bacteria have infected the cells TLR9 is the main initiator of the inflammatory response. This seems to be consistent with the fact that TLR9 is only expressed intracellularly, and thus the bacteria must internalise in order to interact with these receptors.
Interestingly, when we investigated whether Ureaplasma
species could induce the formation of receptor clusters on the cell surface of human amniotic epithelial cells, it was shown that they could induce receptor clusters comprising of TLR2/6 on the cell surface. These clusters seem to form within lipid rafts. Since these pathogens trigger responses via TLR2, it is possible that CD36 could act as a key molecule within the receptor cluster. CD36 has recently been shown to associate with TLR2 
and this interaction might exacerbate the inflammatory response.
Furthermore, when we investigated whether TLR9 was triggering signalling from the endosomes in response to Ureaplasma infection, it was shown that following Ureaplasma infection, TLR9 co-localised in the endosomes with the signalling adaptor molecule MyD88. Therefore, our data suggests that Ureaplasma CpGDNA must be targeted to endosomal compartments upon internalisation. TLR9 is perfectly placed in endosomal compartments in order to be able to “sample” this endocytosed material and trigger cytokine response whenever this molecular “signature” is recognised.
Overall our data suggests that the inflammatory response triggered by Ureaplasma
in amniotic epithelial cells is mediated by the synergic activation of multiple TLRs. It seems that a part of the inflammatory response is triggered by lipoproteins, such as MBA, in the initial stages of the bacterial attachment to the cell surface and is mediated via TLR2. MBA is a surface exposed lipoprotein, which can undergo size and phase variation in vitro
and in vivo
. It has been previously shown that MBA size variation is associated with the severity of chorioamnionitis in a pregnant sheep model of intra-amniotic Ureaplasma
, therefore MBA size variation might influence the interaction of MBA with TLR2/TLR6 heterodimers and thus determine the intensity of the innate immune response and consequently the severity of amniotic fluid infection. One effective microbial strategy for avoiding host recognition is the modification/variation of PAMPs. Since our study demonstrates that MBA is a PAMP, the primary function of this antigenic variation could be to evade the innate immune response, and thus MBA variation could influence the virulence of the different strains. This might explain why certain Ureaplasma
isolates are more associated with severe disease than others.
Although, it has been previously suggested that in addition to TLR2, TLR4 might be activated in response to Ureaplasma
, in our study we only observed TLR2 activation in response to all Ureaplasma
serovars tested. The fact that MBA engages TLR2, and not TLR4, might shed more light into the reasons why these organisms can develop chronic, low-level inflammation of amniotic epithelial cells leading to PROM. Activation via TLR2 results in a subdued inflammatory response 
allowing the organism to establish a chronic foothold in amniotic epithelial cells.
Synergic inflammatory response of TLR2, TLR6 and TLR9 seem to produce a chronic inflammatory response against Ureaplasma, which could eventually lead to irreversible injury of fetal membranes. There are several clinical implications of our findings, since the current study is the first study into the mechanisms by which Ureaplasma species cause chronic inflammation of the amniotic epithelium, and might help us find new TLR-based therapeutic targets for Ureaplasma-induced chorioamniotitis.