To facilitate scalable profiling of single cells, we developed split-pool ligation-based transcriptome sequencing (SPLiT-seq), a single-cell RNA-seq (scRNA-seq) method that labels the cellular origin of RNA through combinatorial barcoding. SPLiT-seq is compatible with fixed cells or nuclei, allows efficient sample multiplexing, and requires no customized equipment. We used SPLiT-seq to analyze 156,049 single-nucleus transcriptomes from postnatal day 2 and 11 mouse brains and spinal cords. More than 100 cell types were identified, with gene expression patterns corresponding to cellular function, regional specificity, and stage of differentiation. Pseudotime analysis revealed transcriptional programs driving four developmental lineages, providing a snapshot of early postnatal development in the murine central nervous system. SPLiT-seq provides a path toward comprehensive single-cell transcriptomic analysis of other similarly complex multicellular systems.

The ability to predict the impact of cis-regulatory sequence on gene expression would facilitate discovery in fundamental and applied biology. Here, we combine polysome profiling of a library of 280,000 randomized 5′ UTRs with deep learning to build a predictive model that relates human 5′ UTR sequence to translation. Together with a genetic algorithm, we use the model to engineer new 5′ UTRs that accurately direct specified levels of ribosome loading, providing the ability to tune sequences for optimal protein expression. We show that the same approach can be extended to chemically modified RNA, an important feature for applications in mRNA therapeutics and synthetic biology. We test 35,000 truncated human 5′ UTRs and 3,577 naturally occurring variants and show that the model predicts ribosome loading of these sequences. Finally, we provide evidence of 45 SNVs associated with human diseases that substantially change ribosome loading and thus may represent a molecular basis for disease. The sequence of the 5′ untranslated region (5′ UTR) is a primary determinant of translation efficiency 1,2. While many cis-regulatory elements within human 5′ UTRs have been characterized individually, the field still lacks a means to accurately predict protein expression from 5′ UTR sequence alone, limiting the ability to estimate the effects of genome-encoded variants and the ability to engineer 5′ UTRs for precise translation control. Massively parallel reporter assays (MPRAs)-methods that assess thousands to millions of sequence variants in a single experiment-coupled with machine learning have proven † Corresponding author. gseelig@uw.edu. Author contributions P.J.S and B.W. designed and performed experiments, performed data analysis and modeling, and wrote the manuscript. D.R. performed fluorescence validation experiments. V.P. and I.M. wrote the manuscript. D.R.M. helped design polysome profiling. G.S. designed experiments and wrote the manuscript.

Ribosomal ribonucleic acid (RNA), transfer RNA and other biological or synthetic RNA polymers can contain nucleotides that have been modified by the addition of chemical groups. Traditional Sanger sequencing methods cannot establish the chemical nature and sequence of these modified-nucleotide containing oligomers. Mass spectrometry (MS) has become the conventional approach for determining the nucleotide composition, modification status and sequence of modified RNAs. Modified RNAs are analyzed by MS using collision-induced dissociation tandem mass spectrometry (CID MS/MS), which produces a complex dataset of oligomeric fragments that must be interpreted to identify and place modified nucleosides within the RNA sequence. Here we report the development of RoboOligo, an interactive software program for the robust analysis of data generated by CID MS/MS of RNA oligomers. There are three main functions of RoboOligo: (i) automated de novo sequencing via the local search paradigm. (ii) Manual sequencing with real-time spectrum labeling and cumulative intensity scoring. (iii) A hybrid approach, coined ‘variable sequencing’, which combines the user intuition of manual sequencing with the high-throughput sampling of automated de novo sequencing.

All tRNAs undergo post-transcriptional chemical modifications as part of their natural maturation pathway. Some modifications, especially those in the anticodon loop, play important functions in translational efficiency and fidelity. Among these, 1-methylguanosine, at position 37 (m 1 G 37 ) of the anticodon loop in several tRNAs, is evolutionarily conserved and participates in translational reading frame maintenance. In eukaryotes, the tRNA methyltransferase TRM5 is responsible for m 1 G formation in nucleus-encoded as well as mitochondria-encoded tRNAs, reflecting the universal importance of this modification for protein synthesis. However, it is not clear what role, if any, mitochondrial TRM5 serves in organisms that do not encode tRNAs in their mitochondrial genomes. These organisms may easily satisfy the m 1 G 37 requirement through their robust mitochondrial tRNA import mechanisms. We have explored this possibility in the parasitic protist Trypanosoma brucei and show that down-regulation of TRM5 by RNAi leads to the expected disappearance of m 1 G 37 , but with surprisingly little effect on cytoplasmic translation. On the contrary, lack of TRM5 causes a marked growth phenotype and a significant decrease in mitochondrial functions, including protein synthesis. These results suggest mitochondrial TRM5 may be needed to mature unmethylated tRNAs that reach the mitochondria and that could pose a problem for translational fidelity. This study also reveals an unexpected lack of import specificity between some fully matured and potentially defective tRNA species.

Establishment of the early genetic code likely required strategies to ensure translational accuracy and inevitably involved tRNA post-transcriptional modifications. One such modification, wybutosine/wyosine is crucial for translational fidelity in Archaea and Eukarya; yet it does not occur in Bacteria and has never been described in mitochondria. Here, we present genetic, molecular and mass spectromery data demonstrating the first example of wyosine in mitochondria, a situation thus far unique to kinetoplastids. We also show that these modifications are important for mitochondrial function, underscoring their biological significance. This work focuses on TyW1, the enzyme required for the most critical step of wyosine biosynthesis. Based on molecular phylogeny, we suggest that the kinetoplastids pathways evolved via gene duplication and acquisition of an FMN-binding domain now prevalent in TyW1 of most eukaryotes. These findings are discussed in the context of the extensive U-insertion RNA editing in trypanosome mitochondria, which may have provided selective pressure for maintenance of mitochondrial wyosine in this lineage.

SUMMARY In cells, tRNAs are synthesized as precursor molecules bearing extra sequences at their 5′ and 3′ ends. Some tRNAs also contain introns which, in archaea and eukaryotes, are cleaved by an evolutionarily conserved endonuclease complex that generates fully functional mature tRNAs. In addition, tRNAs undergo numerous post-transcriptional nucleotide chemical modifications. In Trypanosoma brucei the single intron-containing tRNA (tRNATyrGUA) is responsible for decoding all tyrosine codons; therefore, intron removal is essential for viability. We show by molecular and biochemical approaches the presence of several non-canonical editing events, within the intron of pre-tRNATyrGUA, involving guanosine to adenosine transitions (G to A) and an adenosine to uridine transversion (A to U). The RNA editing described here is required for proper processing of the intron, establishing for the first time the functional significance of non-canonical editing with implications for tRNA processing in the deeply divergent kinetoplastid lineage and eukaryotes in general.

Predicting the impact of cis-regulatory sequence on gene expression is a foundational challenge for biology. We combine polysome profiling of hundreds of thousands of randomized 5′ UTRs with deep learning to build a predictive model that relates human 5′ UTR sequence to translation. Together with a genetic algorithm, we use the model to engineer new 5 UTRs that accurately target specified levels of ribosome loading, providing the ability to tune sequences for optimal protein expression. We show that the same approach can be extended to chemically modified RNA, an important feature for applications in mRNA therapeutics and synthetic biology. We test 35,000 truncated human 5′ UTRs and 3,577 naturally-occurring variants and show that the model accurately predicts ribosome loading of these sequences. Finally, we provide evidence of 47 SNVs associated with human diseases that cause a significant change in ribosome loading and thus a plausible molecular basis for disease.The sequence of the 5′ untranslated region (5′ UTR) is a primary determinant of translation efficiency (1, 2). While many cis-regulatory elements within human 5′ UTRs have been characterized individually, the field still lacks a means to accurately predict protein expression from 5′ UTR sequence alone, limiting the ability to estimate the effects of genome-encoded variants and the ability to engineer 5′ UTRs for precise translation control. Massively parallel reporter assays (MPRAs) -methods that assess thousands to millions of sequence variants in a single experiment -coupled with machine learning have proven useful in addressing similar voids by producing quantitative biological insight that would be difficult to achieve through traditional approaches (3-9).We report the development of an MPRA that measures the translation of hundreds of thousands of randomized 5′ UTRs via polysome profiling and RNA sequencing. We then use the data to train a convolutional neural network (CNN) that can predict ribosome loading from sequence alone. Earlier MPRAs designed to learn aspects of 5′ UTR cis-regulation relied on FACS (10, 11) or growth selection (12) to stratify libraries by activity. These techniques require the expression of a single library variant per cell that must be transcribed within the cell from a DNA template, making it difficult to distinguish between the effects of transcriptional and translational control. Polysome

No abstract