Impacts on the Placenta-Brain Axis in Pregnant Mice Devoid of Gut Microbiota

Gut microbiota play an important role in regulating individual health. It is also increasingly apparent that changes in maternal gut microbiota result can render her offspring more susceptible to diseases later in life. Such bacterial induced changes likely originate in the womb. As the placenta is the primary communication organ between mother and conceptus, it is vulnerable to in utero environmental changes, including those associated with maternal gut microbiota. The placenta also relays nutritional and other factors, including serotonin, to the developing fetal brain that help guide development of this organ. Thus, placental disruptions can influence the placenta-brain axis and subsequently neurobehavioral programming with potential long-term consequences. One mechanism by which changes in maternal gut microbiota might effect the placenta are through alterations in bacterial short chained fatty acids (SCFA). The hypothesis, thus, tested in the current studies is that absence of a maternal gut microbiota, as occurs in germ-free (GF) mice, would impact bacterial SCFA in her fecal samples and in the fetal placenta and brain. Secondarily, we tested whether transcriptomic changes would be evident in the placenta and fetal brain from conceptuses derived from GF relative to multi-pathogen free (MPF) pregnant females. Overall design: RNA was isolated from frozen placenta and fetal brain samples from conceptuses derived from pregnant germ-free (GF) and multi-pathogen free (MPF) females. We previously determined the sex of the conceptus by usage of a PCR based on the Y-chromosome located Sry gene. This gene identified those conceptuses that were male from those who were female. Quality and quantity of RNA was determined by fragment analyzer. RNA was then subjected to RNAseq. Resultant libraries were pooled and run on an Illumina NovaSeq X sequencer with a paired end 100 bp read format on a 10B flow cell to generate ~100 million paired reads per sample. The RNA-seq data analysis pipeline commenced with quality assessment of the raw Illumina sequencing reads using FastQC. High-quality reads were then aligned to the mouse reference genome (GRCm39) by using the Hisat2 aligner with its index-building functionality. Following alignment, transcript assembly and quantification were carried out with Cufflinks, which estimates gene expression levels in terms of FPKM (Fragments Per Kilobase of transcript per Million mapped reads). To identify differentially expressed genes (DEGs), we used Cuffdiff. For both placenta and brain samples, DEGs were defined by using a log2 fold change (log2FC) threshold of 1 and a q-value < 0.05. Visualization of differential expression results is performed using Python. Volcano plots were generated for each comparison with the matplotlib package to highlight significantly upregulated and downregulated genes. Principal Component Analysis (PCA) was conducted on the gene expression profiles (FPKM values) with the scikit-learn library to assess sample clustering, and the resulting components were visualized with seaborn package for clear and informative plotting.