Ureaplasmas are one of the most common microbes isolated from the amniotic fluid of women during mid-trimester, but only a subset of women with positive cultures develop adverse pregnancy outcomes such as PTL, PROM, fetal infection and fetal inflammatory response syndrome 
. The pathogenesis of ureaplasmal associated adverse pregnancy outcomes is further complicated by the fact that most placental infections involving ureaplasmas are polymicrobial 
. However, the impact of mixed bacterial infection on ureaplasmal pathogenesis is not clear since the inflammatory phenotype associated with pure ureaplasmal infections is indistinguishable from mixed bacterial infections that contain Ureaplasma
. Since an integral part of ureaplasmal associated disease involves a robust host inflammatory response, it has been suggested that the host’s reaction to the presence of the microbe is a major determinant in the development of adverse pregnancy outcome 
. By using genetically distinct mouse strains with divergent innate immune responses, we have provided evidence that supports this theory (summarized in ). Despite the fact that both C57BL/6 and BALB/c mice had equivalent placental microbial load, only the BALB/c mouse exhibited a greater proportion of fetal infection. Further, a striking feature in the BALB/c mouse was the presence of extensive neutrophilic infiltrates that were not present in the C57BL/6 mouse. This prevalence of neutrophils within the tissues of BALB/c mice may account for the increased rate of fetal infection along with the greater range of fetal pathology that was observed in this strain.
Predominant BALB/c and C57BL/6 response profile to U. parvum intrauterine infection.
Our experimental approach was designed to elicit both maternal and fetal inflammatory responses and to ensure a high rate of placental and fetal infection. Therefore, a high dose of U. parvum
was deposited into the uterine lumen instead of the amniotic cavity, which allowed for microbial dissemination that resembled ascending infection. As supported by our data, both mouse strains developed high placental infection rates (greater than 90%) that contained microbial loads (CFU of 103
or greater) reported to correlate with intrauterine inflammatory responses such as chorioamnionitis, PROM, fetal morbidity and mortality in humans 
. Immunodetection of U. parvum
demonstrated that the in situ
distribution of the microbe within the placenta, amniotic cavity, and fetal intestine was also the same in both mouse strains. Therefore, the different rates of fetal infection between C57BL/6 and BALB/c mice cannot be attributed to differences in microbial load.
Since our experimental approach was designed to mimic ascending infection, we cannot exclude the possibility that there may have been some variation in the timing of intra-amniotic invasion among the fetal units within each uterine horn. However, our experimental design adjusted for this potential variation in several ways. First, inherent variations were blocked in our analysis by using all fetal-placental tissues within each uterine horn, including those farthest from the inoculation site. Second, studies in our laboratory have demonstrated that the rate of U. parvum spread to all fetal units within the uterus is rapid. For example, a pilot study that included 7 C57BL/6 dams showed that 91% of placentas and 33% of fetuses were colonized with U. parvum by 24 hours post-inoculation (unpublished data). Therefore, analysis of tissues at 72 hours post-inoculation provided sufficient time for adequate microbial dissemination into each feto-placental unit, which was confirmed by a placental colonization rate of 100% in C57BL/6 mice. Interestingly, we observed a higher rate of fetal infection in our pilot study (45% at 24 hours vs. 18% at 72 hours), which implies that some microbial clearance may be occurring in C57BL/6 mice as the duration of infection increases.
A striking feature in BALB/c dams was the extensive neutrophilic infiltration that was present in both uterine and placental tissues. Most neutrophils in BALB/c tissues appeared to be undergoing necrotic death instead of apoptosis. Necrotic death prolongs neutrophil activation and inflammation, delaying neutrophil clearance and the reparative response 
. Other features that support ongoing inflammation in BALB/c tissues included increased expression of pro-inflammatory mediators, thickening of the amniotic basement membrane, and extensive desquamation of epithelial cells. Taken together, these features are consistent with prolonged inflammation of at least 36 to 48 hours and with breach of the maternal-fetal barrier 
, which may have contributed to the increased rate of fetal infection observed in this mouse strain.
The divergent placental inflammatory responses that were observed in C57BL/6 and BALB/c mice may be due to inherent differences in their neutrophil function, toll-like receptor (TLR) responsiveness 
, and/or macrophage dominant profiles 
. These host specific factors can impact disease outcome independently of U. parvum
mediated effects on the host. For example, Allenbach et al 
demonstrated that C57BL/6 neutrophils undergo extrinsic apoptosis during the early phase of L. major
infection, which is important for reducing the damaging effects of inflammation during infection with this pathogen. In our study we observed scant influxes of neutrophils in U. parvum
infected uterine and placental tissues of C57BL/6 mice without evidence of extensive tissue necrosis, which may be due to non-reactive neutrophil clearance 
in this mouse strain. Further, C57BL/6 mice exhibited higher levels of placental IL-10 production that did not significantly dissipate in the presence of infection. This would also support resolution of inflammation and a reparative response 
. Conversely, BALB/c mice infected with L. major
have defective extrinsic neutrophil apoptosis coupled with severe protracted neutrophilic inflammation 
, which is similar to what we observed in U. parvum
infected tissues of this mouse strain. BALB/c mice also have decreased responsiveness to TLR agonists or cytokines 
, and typically produce a suppressive or macrophage M2 dominant immune response 
. Specifically, macrophages from BALB/c mice typically produce prostaglandin E2
) when exposed to lipopolysaccaride, which inhibits a TH1 cytokine response in these animals 
. BALB/c mice also produce less TNF-α and IL-12 in response to MALP-2, which is a mycoplasma protein that is a TLR2/6 specific agonist 
. Compared to C57BL/6 mice, placental tissues of sham inoculated BALB/c mice exhibited decreased expression of IL-1β, IL-6, and TNF-α, which is more in tune with M2 macrophage polarization 
. This feature may be particularly critical in adverse pregnancy outcomes involving ureaplasmas since ureaplasmas elicit PGE2,
IL-1β, TNF-α, and IL-10 from human amniochorion and choriodecidual explants 
, which is neither characteristic of a TH1 or TH2 cytokine response. Thus, U. parvum
infection in the M2 dominant BALB/c mouse may trigger an attenuated innate response that is not microbicidal 
, but instead results in a detrimental inflammatory response.
The fetal inflammatory response profiles observed in C57BL/6 and BALB/c mice also implied that the extent of fetal infection was different in these strains. For example, pathology was only detected in the liver and intestine of C57BL/6 fetuses, and U. parvum
was only identified in the intestinal lumen of these animals by IFA. This would suggest that infection was localized and most likely occurred by fetal ingestion of contaminated amniotic fluid. On the other hand, the presence of varying degrees of chorionic vasculitis and/or umbilical vasculitis 
, extensive influx of neutrophils into the amniotic cavity, coupled with multiple fetal organ pathology (brain, lung, heart, liver and intestine) in BALB/c fetuses was more consistent with fetal sepsis and fetal inflammatory response syndrome 
. Other indicators of fetal sepsis and/or fetal inflammatory response syndrome were elevated IL-6 and S100A8 in placental tissues 
. Although hepatic lesions observed in U. parvum
infected fetuses is not a feature of fetal inflammatory response syndrome, hepatic lesions have been observed during human fetal infection with Ureaplasma
as well as in the sheep model of U. parvum
intra-amniotic infection 
Limitations with this study include 1) lack of earlier time points that would elucidate the temporal progression of maternal and fetal inflammatory responses during microbial spread into the amniotic cavity, 2) clear distinction of fetal leukocytes from maternal leukocytes within the amniotic cavity, and 3) the lack of fetal organ specific cultures that would have clarified if BALB/c fetuses indeed had multiple organ infection and sepsis. Despite these limitations, this study demonstrated that the host genetic background significantly impacts disease outcome during intrauterine infection with U. parvum. Although our study cannot conclude that similar effects may be observed with intrauterine infections caused by other microbial species, it does underscore the need to consider the host genetic background as a factor in disease pathogenesis, especially studies that use the mouse as an infection model.