LAUSR

Uterine microbiota: are there differences within three major time points of the menstrual cycle in healthy controls, recurrent miscarriage (RM) and recurrent implantation failure (RIF) patients? Compared to controls, RM and RIF patients showed an altered uterine microbiota throughout the menstrual cycle, with a lower dominance of lactobacilli. In contrast to the former notion of a sterile womb, bacterial colonization in the uterus and the placenta has been demonstrated. Studies showed that Lactobacillus-dominated endometrial microbiota correlate with reproductive success. Moreover, the presence of non-Lactobacillus-dominated microbiota, especially with detection of Gardnerella and Streptococcus in the endometrial fluid, seems to be associated with lower implantation-, ongoing pregnancy- and live birth-rates. However, intra-cycle variations in healthy women as well as possible alterations in patients with RM or RIF remain unknown. In total, n = 20 RM patients (≥ 3 consecutive miscarriages), n = 20 RIF patients (≥3 fresh or frozen embryo transfers with negative serum hCG, good quality embryos) and n = 10 healthy controls (no pregnancy) were included in this study. All patients had a 28 day menstrual cycle. During follicular, ovulatory and luteal-phase, after a thorough cleaning of the cervix, a flexible catheter was introduced into the uterine cavity and a uterine flushing with 1ml of NaCl was performed. Bacterial DNA was extracted using a QIAamp DNA kit (Qiagen) in combination with a PrecellysR24 homogenizer (Peqlab, Erlangen, Germany) according to the manufacturer’s instructions. The V3-V4 region of the bacterial 16S rRNA gene was amplified. Samples were pooled in equimolar ratios and progressed to pyrosequencing using an Illumina MiSeq se-quencer with MiSeq Kit V2 (250 bp paired-end). Analysis of 16S rRNA data, including alpha- and beta-diversity, were calculated using the phyloseq package in R. For the Shannon index (species richness and evenness) a significant decrease during the ovulation period was shown in the control group, indicating a more uniform microbiota (p < 0.05). This loss of diversity was not shown in RIF and RM patients. Overall, we could observe a higher similarity in taxonomic distribution in RM compared to the RIF patients. Longitudinal dynamics included increases in Firmicutes (CTRL and RM only) and a concomitant loss of Proteobacteria. Notably, significant amounts of bacteroides were only detected in the RIF patients. Actinobacteria were more frequent in both, RM and RIF as compared to controls. To minimize the impact of a potential contamination, we performed pre-experiments with paired samples both from the vaginal fornix and the endometrial cavum and could show a significant difference in overall microbiome configuration. However, the route of sample can still be prone to contamination. For the first time, we were able to show cycle-dependent alterations in the endometrial microbiome. These findings underline the role of an altered endometrial microbiome as a cause for RM and RIF and can contribute to the future establishment of therapeutic strategies in cases of a dysbalanced microbiome. Drks00020803