AIR POLLUTION EXPOSURE IN PREGNANCY: NASAL MICROBIOTA AND EXTRACELLULAR VESICLE COMMUNICATION AS POTENTIAL MECHANISM TO EXPLAIN ADVERSE BIRTH OUTCOMES

ABSTRACT BACKGROUND: Particulate matter (PM) exposure has been linked to the exacerbation of respiratory and cardiovascular conditions as well as to adverse effects on fetal growth. To link the cross-talk that might occur between respiratory system and placenta after PM exposure, it has been proposed a novel mechanism of cell to cell communication mediated by extracellular vesicles (EVs). EVs are involved in both biological and pathological processes including pregnancy state. As PM interacts firstly with the nares, bacterial nasal microbiota (bNM) is one of the first compartments hit by PM exposure. This interaction might lead to structural and functional modifications within the bNM, which could cause variations within the EVs signaling network, which might lead to an improper immune response to PM stimuli. AIM: The main aim of the project was to identify how PM exposure might modify the homeostasis and composition of whole EV signaling network and bNM and the leading to a possible impact on newborn development. Subject recruitment: 518 volunteer pregnant women were enrolled during the 11th week of pregnancy at the ‘Clinica Mangiagalli’-Fondazione IRCCS Ca’ Granda – Ospedale Maggiore Policlinico, Milan, Italy. Among them, a group of subjects composed by 65 pregnant women, who agreed to participate to a more complex study protocol, was also identified. Exposure assessment, EV and Microbiota measurement: Exposure to PM concentrations was assessed using data obtained from FARM models for the whole population. In addition, individual exposure to short-term PM levels was retrieved through a personal sampler worn by the subgroup of 65 pregnant women. Plasmatic concentration and cellular origin of EVs were characterized by Nanoparticle tracking analysis (NTA) and flow cytometry, respectively. We investigated the bNM structure and characteristics of 65 pregnant women both at the enrolment (T0) and the following Monday during the cardiovascular screening (T1) through metabarcoding analysis of the V3–V4 regions of the 16s rRNA gene. Statistical analysis: Multivariable linear regression models were applied to test the associations between PM exposure (retrieved from both FARM models and personal sampler) and the majority of the collected outcomes such as maternal, foetal/newborns, and cardiovascular parameters as well as for bNM data. On the other hand, to evaluate possible associations between PM concentrations and EV characteristics negative binomial regression models for count data with over-dispersion were performed. In addition, multiple comparison method based on Benjamini-Hochberg False Discovery Rate (FDR) were applied for high number of comparisons. RESULTS In the whole population, PM10 exposure (measured at different time windows) resulted in decreased release of total amount of EVs, with the strongest effect related to concentration measured 13 weeks before the enrolment (13wks), whereas an inverse tendency was observed for exposure to PM2.5, although these associations were not significant. More reliable data on the finest fractions (PM1, PM2.5 and PM4) are given by personal sampler worn by a subgroup of women for a very short time period preceding the blood drawing (1.5 hours). As we considered this extremely acute effect, we observed a generalized increment in the EV count. Noteworthy, among the different analyzed EV subtypes, the levels of HERV-w+ EVs were the only to be increased by each tested PM10 time-lag. The same models applied on bNM data showed a reduction in terms of diversity (Shannon/Faith_pd ratio) and relative abundance of the genera Corynebacterium spp. and Staphylococcus spp. In addition, when the possible role of bNM as effect modifier between PM exposure and EVs release was investigated, we observed for the pregnant women with a balanced bNM an increment in terms of circulating EVs after daily PM stimuli. Moreover, increments of the heart rate values were observed after exposure to both PM10 and PM2.5 levels measured the day before the cardiovascular screening (Day -1). Focusing on newborn’s outcomes, decrements of the gestational age at birth were associated to PM concentrations measured throughout the gestation or during the 2nd trimester. CONCLUSIONS: To our knowledge, this is the first exploring the role of bNM and the EV cross-talk in determining the effects of PM exposure levels on healthy pregnancies as well as on newborn outcomes. The results obtained so far might suggest a possible role exerted by both EV concentration and the bNM in pregnant women in mediating the effects of PM exposure.