Buffering of transcription rate by mRNA half-life is a conserved feature of Rett syndrome models

Abstract

Models of MECP2 dysfunction in Rett syndrome (RTT) assume that transcription rate changes directly correlate with altered steady-state mRNA levels. However, limited evidence suggests that transcription rate changes are buffered by poorly understood compensatory post-transcriptional mechanisms. Here we measure transcription rate and mRNA half-life changes in RTT patient neurons using RATE-seq, and reinterpret nuclear and whole-cell RNAseq from Mecp2 mice. Genes are dysregulated by changing transcription rate only or half-life only and are buffered when both are changed. We utilized classifier models to understand the direction of transcription rate changes based on gene-body DNA sequence, and combined frequencies of three dinucleotides were better predictors than contributions by CA and CG. MicroRNA and RNA-Binding Protein (RBP) motifs were enriched in 3’UTRs of genes with half-life changes. Motifs for nuclear localized RBPs were enriched on buffered genes with increased transcription rate. Our findings identify post-transcriptional mechanisms in humans and mice that alter half-life only or buffer transcription rate changes when a transcriptional modulator gene is mutated in a neurodevelopmental disorder.

Transcriptional buffering and 3ʹUTR lengthening are shaped during human neurodevelopment by shifts in mRNA stability and microRNA load

“…Given the consistently high frequency of buffered, boosted and half-life only gene sets in iPSC, NPC and Neu, it is highly likely that equivalent modes of gene regulation will be found when differentiating iPSC into other human cell types to model development and disease. Importantly, the same modes of gene regulation should be widely found in vivo, as we have already verified buffered and other gene regulation modes in a high-confidence RNA-seq resource of control and Mecp2 null mouse brain samples 10 .…”

“…Reassuringly, pluripotent stem cell (miR-302, miR-200) 39,40 and neuron (miR-137, miR-7) 41 specific miRNAs accumulated in the appropriate cell types after both transitions (Fig 4B ) (Supplementary table 3). Small RNA-seq results of the same control neuron samples have already been reported in comparison to isogenic Rett syndrome neurons 10 . As expected, we observed a remodeling of the miRNA landscape with 676 (25%) and 585 (22%) miRNAs changing steadystate abundance after iPSC to NPC and NPC to Neu transitions, respectively (Fig 4B ).…”

“…To calculate relative and absolute differences in the miRNA population in iPSC, NPC and Neurons, small RNAs were extracted from two replicates of both lines using the same number of cells followed by the addition of a set of spike-in RNAs exactly as described 10 . Small RNAs were extracted from 500,000 cells of each line using the SPLIT RNA extraction Kit (Lexogen).…”

“…RNA-seq libraries were prepared for each time-point and steady-state sample exactly as described 10 . In brief, the QuantSeq 3ʹ mRNA-Seq Library Prep Kit FWD for Illumina (Lexogen) was automated on the NGS WorkStation (Agilent) at The Centre for Applied Genomics (TCAG).…”

“…Processing was performed exactly as described 10 . In brief we trimmed reads in 4 steps using cutadapt version 1.10 50 .…”