LAUSR

Formalin-Fixed Paraffin-Embedded (FFPE) tissues are routinely collected, archived, and used for clinical diagnosis, including maternal and neonatal health. Applying FFPE samples to microbiota research would be beneficial to reduce preparation, storage and costs associated with limited available frozen samples. This research aims to understand if FFPE fetal membrane samples are comparable to frozen tissues, which are the current gold standard for DNA microbiota analysis. Extracted DNA from nine matched paired patients were sequenced by Illumina sequencing of the V4 16S rRNA gene region. This included duplicate frozen amnion and chorion fetal membrane rolls or FFPE combined amniochorionic samples. Negative controls of surrounding wax blocks and DNA extraction reagents were processed alongside samples using identical methods. DNA quality and quantity was assessed by NanoDrop, agarose gel electrophoresis and Bioanalyzer. Decontam and SourceTracker were integrated into microbiota analysis to identify the presence of contaminating sources. The bacterial profile and nine genera differed between FFPE and frozen fetal membranes. There were no differences in bacterial profiles between FFPE samples and corresponding wax negative controls, with 49% of bacteria in FFPE fetal membrane samples matched to the source origin of paraffin wax, and 40% originating from DNA extraction reagent sources. FFPE samples displayed high fragmentation and low quantity of extracted DNA compared to frozen samples. The microbiota of FFPE fetal membrane samples is influenced by processing methods, with the inability to differentiate between the microbiota of the tissue sample and the surrounding wax block. Illumina sequencing results of FFPE and frozen fetal membrane samples should not be compared using the methods employed here. Variation could be influenced by limitations including storage time, DNA extraction and purification methods. To utilise FFPE fetal membrane samples in microbiota research then contamination prevention and detection methods must be included into optimised and standardised protocols, with recommendations presented here.