It is estimated globally that approximately 12 million babies are born pre-term each year, making the prevention of pre-term birth (PTB) one of the highest priorities for international obstetric research (1). Intrauterine infection (IUI) and subsequent inflammation of the extra-placental membranes (chorioamnionitis) accounts for approximately 40% of all spontaneous PTBs and is the major cause of early PTB (<32 weeks of gestation). IUI is typically “silent” (undiagnosed) until the onset of pre-term labor at which point it is often too late for treatment as chorioamnionitis is well established, the risk of fetal inflammatory response syndrome (FIRS) is high (2), and tocolysis is ineffective. Identifying women at risk of infection-associated PTB sufficiently early in pregnancy to allow therapeutic intervention would be a significant advance in the prevention of PTB.

Traditional thinking associates IUI with the ascension of bacteria from cervicovaginal fluid, resulting in intra-amniotic infection and immune stimulation within the otherwise sterile intrauterine environment (3). It is now clear, however, that the placenta and extra-placental membranes can no longer be considered strictly sterile (4, 5). Instead, they are home to a unique microbiome of non-pathogenic commensals; the presence of which is normal and not associated with early delivery or adverse pregnancy outcomes (4, 6). Histological and immunological analysis of intrauterine tissues and fluids suggests that the nature and magnitude of inflammatory response associated with bacterial colonization may be key in determining obstetric outcome (6–8).

Recent placental microbiological studies have reignited debate regarding the role of differential virulence (9), poly-microbial interactions (10), host genetics (11), and immune factors (12) in determining obstetric outcome. While attention has primarily been placed on defining the local immune response to bacteria within placental tissues, the role and significance of the maternal systemic immune response to the infection has been largely neglected. Yet, studies conducted at the end of the last century strongly suggested that the maternal immune response to commensal microorganisms found in the urogenital tract in pregnancy – in particular the Ureaplasma species – may provide us with important clues as to why some women are at risk of adverse pregnancy outcomes while the majority are not.

In this mini-review, we discuss in detail the somewhat contradictory evidence relating to the presence and nature of maternal antibodies to Ureaplasma sp. and their significance in determining and predicting obstetric outcome. A specific focus is placed on the potential clinical utility of serological analysis in the identification of women at elevated risk of PTB.

Ureaplasma spp. are generally considered commensal microorganisms (13–16) and are classified into two species and 14 distinct serovars (SV). SV1, SV3, SV6, and SV14 belong to U. parvum species and the remaining ten SV to U. urealyticum (17). Ureaplasma sp. commonly colonize the urogenital tract of both males and females (18, 19). Vaginal colonization rates can vary greatly in non-pregnant women (up to 70%) (20, 21) and women with uncomplicated pregnancies [2.7–70%; reviewed in Ref. (22)]. Ureaplasma spp. are some of the most frequently identified microorganisms in placental tissues and amniotic fluids (AF) from pre-term deliveries (23–25). Colonization of the placenta with Ureaplasma sp. has been demonstrated to be an independent risk factor for chorioamnionitis [odds ratio (OR), 11.27; 95% CI, 5.09–24.98] (7). Detection rates in AF vary from 0% to 19% in early mid-trimester (26–28), to 2–80% at pre-term labor (26, 29) and 18–100% with pre-labor premature rupture of membranes (PPROM) (26, 30). A meta-analysis of 22 studies found a significant association between the presence of Ureaplasma sp. in the vagina and AF with PTB (22). However, Ureaplasma sp. colonization in the urogenital tract and AF is a relatively common finding in pregnant women and alone it is not sufficiently predictive of PTB to be clinically useful (8).

The reasons why commensal Ureaplasma sp. cause ascending IUI leading to PTB in only a subset of women are still unknown but are likely to be complex and multifactorial (Figure 1). SV-specific virulence has been proposed as an important determinant of risk of adverse outcome. Of the two Ureaplasma species, U. parvum is the most commonly isolated species from clinical samples (31), with two of its SV – SV3 and SV6 – associated with worse pregnancy outcomes (32–34). Unfortunately, most studies of IUI do not differentiate between Ureaplasma SV so data on the relationship between SV prevalence and risk are lacking. Poly-microbial interactions may also be significant as diagnosis of abnormal bacterial flora or bacterial vaginosis (BV) (i.e., a high Nugent score) is also a risk factor for PTB (35), independent of Ureaplasma sp. colonization status. It is worth noting that poly-microbial colonization of the amniotic cavity is common in pre-term deliveries, with around half of all infected AF containing two or more microorganisms (23). Co-colonization of the vagina with a Ureaplasma sp. and another genital Mycoplasma sp. has been reported to be associated with more severe adverse pregnancy outcomes compared to colonization with a single organism (36). Competence of the cervical barrier to microbial ascension is also likely to play a role in determining risk of PTB. A short cervix in pregnancy has been associated with increased risks of microbial colonization (37) and spontaneous PTB (38, 39).

Finally, maternal/fetal immunological tolerance or competence is a likely determinant of obstetric outcome associated with colonization by Ureaplasma sp. and other microorganisms. Surface-exposed lipoproteins of Ureaplasma sp. activate the pro-inflammatory transcription factor NF-κB through TLR ligation (40, 41), although exposure of intrauterine tissues to Ureaplasma sp. does not generally trigger a robust inflammatory response. Nevertheless, intra-amniotic infection with U. parvum has been causally linked to chorioamnionitis, FIRS and PTB in animal models (42–44). This is consistent with observations from clinical studies showing that chorioamnionitis/funisitis is more likely in pregnancies infected with Ureaplasma sp. as opposed to other bacteria (7). On the other hand, most women with vaginal colonization and local inflammatory response will not deliver preterm (24, 45). Combs, Gravett (8), proposed a model of IUI comprising of five subgroups: (1) microbial colonization plus inflammation (AF IL-6 >11.3 ng/ml); (2) severe inflammation with no detectable microorganisms (AF IL-6 >11.3 ng/ml); (3) mild inflammation with no detectable microorganisms (AF IL-6 of 2.6–11.2 ng/ml); (4) microbial colonization with no inflammation; and (5) absence of infection or inflammation. This more complicated scenario highlights the need to accurately stratify women at risk of PTB associated with IUI.

Ureaplasma sp. have been shown to stimulate pro-inflammatory responses in fetal membranes (73), choriodecidual explants (74), and preterm and term cord blood (75). However, studies of white blood cell responses to Ureaplasma sp. in pregnancy have not proven to be useful to date, with no differences in responsiveness observed in leukocytes from women at risk of PTB compared to those with normal pregnancies (12). Reyes et al. explained differential pathogenesis of Ureaplasma sp. in urinary tract infections (UTIs) by demonstrating the presence of two distinct immune cell profiles (76). Asymptomatic UTIs were characterized by minimal monocytic and lymphocytic infiltration, less tissue damage and increased IFN-γ, while complicated UTIs were associated with greater concentrations of pro-inflammatory cytokines, extensive inflammation and predominantly a neutrophilic response. No comparable data from pregnant women have been published to date. Recently, unconventional lipid antigens such as those present in Ureaplasma sp. (9, 57), and microbe-derived vitamin B metabolites (77), have also been shown to stimulate T cell responses vital for microbial clearance (78, 79). However, the role of these unconventional T cells in the response to Ureaplasma sp. in the context of pregnancy immunity and PTB risk remains unknown and the role of the systemic immune cell responses to Ureaplasma sp. exposure in pregnancy remains relatively unexplored.

Despite advances in our understanding of the causal relationships between intrauterine infection and PTB, major research questions remain unanswered: is there a subpopulation of patients in which Ureaplasma sp. colonization correlates with worse disease outcome? Does the site of colonization determine disease outcome? Are there pathogenic and non-pathogenic SV determining disease outcome? Is there a specific aspect of the maternal immune response to Ureaplasma sp. infection that influences risk of PTB?

Research on specific Ureaplasma sp. SV and virulence factors have failed to yield conclusive results. Despite the earlier clinical promise of antibody-based predictive approaches (26, 62), the assessment of maternal or fetal seropositivity to identify women for prophylactic treatment has not made it into clinical practice. It is unclear when and where exposure to Ureaplasma sp. antigens takes place in pregnancy and when/how commensal colonization becomes an infection. As such, it is difficult to define the role of antibodies in pathogenesis. Antibody cross-reactivity among the Ureaplasma SV and lack of stringent proof of epitope specificity have also limited attempts to use an individual’s antibody response to make a SV-specific diagnosis of infection or predict an outcome.

Before an antibody-based immunological test can be considered as part of a routine antenatal screen, future studies must address technical and scientific issues surrounding the detection and antigen characterization of antibodies to Ureaplasma sp. (41, 46, 47, 54, 58). Key issues to be addressed include: (i) need for the identification of the colonizing SV; (ii) requirement for the detection of a SV-specific or global antibody response; (iii) characterization of immunodominant epitopes; (iv) defining the difference between commensal colonization and infection; (v) determining when and where Ureaplasma sp. are first immunologically detected; and (vi) characterizing the kinetics and magnitude of the antibody response in relation to pregnancy outcome. Based on current knowledge, it seems that the detection and measurement of uncharacterized maternal antibodies against Ureaplasma sp. has limited predictive value for identifying women at elevated risk of infection-driven PTB.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.