A growing body of data exists, providing a causal association between infection of the amniotic cavity, in utero inflammation and preterm birth.10,31–33
species (notably U parvum
) are among the microorganisms most commonly identified in association with preterm birth.16,33–35
Due to the observation that a number of the microorganisms implicated in the etiology of preterm birth are fastidious with respect to nutrient and atmospheric culture conditions, there potentially exists a negative culture bias in the data relating to preterm birth and infection.36
Indeed, there is growing evidence that the utilization of dual culture—PCR-based screening methodologies for amniotic infections—may provide a more accurate picture of infection status than culture alone.36
Irrespective, a number of clinical cohort studies in preterm populations have identified significant associations between positive UP cultures and chorioamnionitis.35
Cases of spontaneous preterm labor were significantly more likely to be culture positive for UP than preterm delivery resulting from medically indicated intervention. Using a molecular approach, Kasper and colleagues recently reported a positive correlation between UP bacterial load and intrauterine inflammation in preterm infants.37
A number of investigators, ourselves included, have employed animal models to demonstrate an association between in utero infection with UP and fetal/uterine inflammation.38–40
Substantive data exist to provide a strong clinical and experimental link between inflammation and preterm birth12,32,34
; however, despite these advances in our understanding, significant questions remain in relation to the relative spatial and temporal contributions of fetal and maternal tissues involved in the initiation and propagation of in utero inflammation.
Adult human keratinocytes express a wide array of Toll-like receptors (TLRs) including TLR-1, TLR-2, and TLR-6, which are presently believed to mediate the innate immune system’s response to UP.41
We assessed the ability of primary fetal ovine keratinocytes (the major cellular constituent of the developing epidermis) to respond to the presence of UP over a 12-hour time course. Analysis of mRNA expression demonstrated that exposure to UP increased expression of MCP-1 between 2 and 6 hours postexposure. In striking contrast to our earlier in vitro studies using E coli
LPS, no significant increases were detected for IL-1β, IL-6, IL-8, TNF-α, or IL-10 (). Monocyte chemoattractant protein 1 (alternatively named CCL-2) is a chemokine demonstrated to possess a wide range of biological and immunomodulatory functions, the most well known of which is as a potent monocyte attractant.42,43
It is expressed by numerous tissues (including endothelial and epithelial cells) in response to cytokines (IL-1, IL-4, and TNF-α) and by exposure to bacterial agonist. Monocyte chemoattractant protein 1 is also involved in numerous biological processes including the promotion of tissue angiogenesis, bone remodelling, and has also been shown to act greatly to increase the permeability of the blood–brain barrier.42
Having first demonstrated that primary fetal ovine keratinocytes respond to UP in a manner substantially different from that induced by E coli
LPS exposure, we sought to assess the nature of the response of the developing fetal skin to infection within the amniotic cavity with UP. In concordance with clinical observations, we have demonstrated previously that the ovine fetal skin responds vigorously to direct in utero exposure to E coli
LPS, resulting in the epidermal migration of immunocytes and upregulation of key proinflammatory cytokines.25
In the present study, we demonstrate that exposure of the developing fetal skin to intra-amniotic UP infection leads to the development of a proinflammatory response involving upregulation of specific cytokines and chemokines, in conjunction with a vigorous and sustained basophilic migration to both the dermis and the epidermis (). The inflammatory response to the 7-day UP exposure was characterized by marked increases in MCP-1, TNF-α, and IL-10. Cytokine/chemokine expression in response to a 69-day UP exposure was significantly reduced in comparison to the 7-day UP-exposure group, with none of the transcripts assessed remaining increased relative to control. We have demonstrated previously that exposure of fetal skin to E coli
LPS induces a striking basophilic infiltration. Data presented here demonstrate a prolonged basophilic infiltration of the epidermis and dermis in fetal skin exposed to UP for either 7 or 69 days ().
Our data suggest that significant differences exist in the mechanistic induction of inflammation in the fetal epidermis following exposure to UP or E coli. Inflammation resulting from E coli LPS exposure appears to involve the expression of proinflammatory cytokines by the epidermal keratinocytes themselves, whereas inflammation resulting from Ureaplasma exposure may involve invading immunocytes in response to elevated MCP-1 signaling from stimulated keratinocytes in the fetal skin.
We have demonstrated previously that prolonged or repeated exposure to E coli
LPS induces endotoxin tolerance in fetal tissues and immunocytes.18,44
This tolerance conveys reduced responses to bacterial agonist and accordingly has significant implications for the preterm infant with respect to immunocompetence and ability to combat nosocomial infections.18,44
Intriguingly, skin from fetuses testing positive for UP after 69 days exposure exhibited significantly lower inflammation that are seen on the skin from 7-day exposures—despite our qPCR analysis demonstrating an equivalent UP gDNA load in each tissue group. These data suggest that the fetal skin develops an immunological tolerance for UP. Recent studies by Peebles and coworkers have suggested that in utero infections caused by multiple microorganisms may be significantly more common than previously thought.45
It will be of great interest to investigate the effect/effects of polymicrobial in utero infection on the ability of the skin and adaptive immune system to respond to the presence of bacterial agonist.
The findings of this study have a number of clinical implications. The fetal skin comprises a significant percentage of total fetal mass and possesses an ample vascular supply and thus has the potential to function as a large source of inflammatory mediators. Although we have demonstrated the fetal skin possess the ability to generate a localized inflammatory response to both E coli LPS and UP, the systemic effects of this inflammation on fetal well-being and development are presently unknown. Additionally, the immature fetal epidermis lacks an outer cornified envelope, potentially allowing inflammatory mediators generated within the fetal skin to cross into the amniotic fluid and interact with other fetal and maternal tissues.
A number of adult inflammatory disorders (ulcerative colitis, arthritis, and cardiovascular disease) are closely associated with the prolonged and inappropriate systemic elevation of inflammatory cytokines, including a number of those shown to be elevated in the present study (TNF-α and MCP-1). Ureaplasma spp are among a number of microorganisms identified in association with preterm birth and there is increasing evidence to suggest that polymicrobial infections of the amniotic fluid are more commonly associated with preterm birth than previously suspected. From a therapeutic perspective, understanding the mechanisms by which microorganisms commonly associated with preterm birth induce in utero inflammation allows for the identification of potential signalling cascade targets which may be able to be manipulated in an attempt to ameliorate in utero inflammation.
The present study has several limitations which need to be taken into consideration; the ability of the developing fetal ovine skin to respond to microorganisms at differing GAs is unknown and potential changes may occur in the viability of UP in utero during a chronic infection. Both factors may impact our analysis of fetal skin exposed to UP for 7 days or 69 days and suggest further investigations in this area are warranted.