Combined single-cell profiling of expression and DNA methylation reveals splicing regulation and heterogeneity

Abstract

BackgroundAlternative splicing is a key regulatory mechanism in eukaryotic cells and increases the effective number of functionally distinct gene products. Using bulk RNA sequencing, splicing variation has been studied across human tissues and in genetically diverse populations. This has identified disease-relevant splicing events, as well as associations between splicing and genomic features, including sequence composition and conservation. However, variability in splicing between single cells from the same tissue or cell type and its determinants remains poorly understood.ResultsWe applied parallel DNA methylation and transcriptome sequencing to differentiating human induced pluripotent stem cells to characterize splicing variation (exon skipping) and its determinants. Our results show that variation in single-cell splicing can be accurately predicted based on local sequence composition and genomic features. We observe moderate but consistent contributions from local DNA methylation profiles to splicing variation across cells. A combined model that is built based on genomic features as well as DNA methylation information accurately predicts different splicing modes of individual cassette exons. These categories include the conventional inclusion and exclusion patterns, but also more subtle modes of cell-to-cell variation in splicing. Finally, we identified and characterized associations between DNA methylation and splicing changes during cell differentiation.ConclusionsOur study yields new insights into alternative splicing at the single-cell level and reveals a previously underappreciated link between DNA methylation variation and splicing.Electronic supplementary materialThe online version of this article (10.1186/s13059-019-1644-0) contains supplementary material, which is available to authorized users.

CD44 alternative splicing is a sensor of intragenic DNA methylation in tumors

“…An important conclusion from this study is that DNA methylation, that at best was considered as a way of fine-tuning alternative splicing, may for small number of genes, have an impact such that the level of DNA methylation present inside the body of these genes can have predictive value when estimating the splicing isoforms produced in a given cell type. This is further supported by a recent study showing that splicing variations at the single cell level can be accurately predicted based on local DNA methylation (Linker et al, 2019). Another intriguing possibility, and perhaps a more interesting one, is that these predicTable genes may serve as sensors of DNA methylation to adapt gene expression to external cues (Fig.…”

Coverage-dependent bias creates the appearance of binary splicing in single cells

“…This unexpected result contrasted with previous single molecule imaging studies of several alternative exons that showed that cell-to-cell variability is minimized and tightly regulated by the splicing machinery in single cells [10]. This led to investigations of the mechanisms that might be responsible for stochastic splicing variability among apparently homogeneous cells, such as variation in DNA methylation [11].…”

“…We used our simulator to investigate how the observed inclusion (Ψ) of cassette exons differs from the underlying Ψ, setting the average capture rate to 10% and the other technical parameters to values that are characteristic of Smart-seq2 datasets (see methods). We considered either a bimodal and binary regime of Ψ (i.e., both isoforms are expressed in the population, but rarely by the same cell; Figure 2b), or a non-binary regime (cells tend to express both isoforms; Figure 2c) [5,11,21]. We simulated the splicing of cassette exons in 500 genes, in a population of 300 single cells.…”

“…A recent study advised that the high dropout rate of scRNAseq makes it fundamentally unsuitable for measuring changes in splicing [14]. Others have implemented workarounds, e.g., using sequence features to predict splicing outcomes in lieu of sufficient sequencing coverage [6], or attempting to identify excess variance beyond technical noise [3,11]. These studies have identified true examples of differential splicing in single cells, but they fundamentally do not explain how scRNA-seq limitations have caused qualitative, not just quantitative, distortions in our understanding of alternative splicing.…”

“…Our simulations make it clear that the reliability of splicing measurements is a function of the initial amount of mRNA, the efficiency of its recovery, the underlying splicing rate, and further distortions from PCR amplification of cDNA. These effects should be considered when interpreting previous studies that used qualitative changes in the observed distribution of the splicing rates [5] or changes in their variance [11] as evidence for regulation of alternative splicing. Considering the true amount of information available for a cassette exon can allow for accurate observations of alternative splicing.…”