Depletion of Hemoglobin Transcripts and Long-Read Sequencing Improves the Transcriptome Annotation of the Polar Bear (Ursus maritimus)

Biochemical-free enrichment or depletion of RNA classes in real-time during direct RNA sequencing with RISER

The heterogeneous composition of cellular transcriptomes poses a major challenge for detecting weakly expressed RNA classes, as they can be obscured by abundant RNAs. Although biochemical protocols can enrich or deplete specified RNAs, they are time-consuming, expensive and can compromise RNA integrity. Here we introduce RISER, a biochemical-free technology for the real-time enrichment or depletion of RNA classes. RISER performs selective rejection of molecules during direct RNA sequencing by identifying RNA classes directly from nanopore signals with deep learning and communicating with the sequencing hardware in real time. By targeting the dominant messenger and mitochondrial RNA classes for depletion, RISER reduces their respective read counts by more than 85%, resulting in an increase in sequencing depth of 47% on average for long non-coding RNAs. We also apply RISER for the depletion of globin mRNA in whole blood, achieving a decrease in globin reads by more than 90% as well as an increase in non-globin reads by 16% on average. Furthermore, using a GPU or a CPU, RISER is faster than GPU-accelerated basecalling and mapping. RISER's modular and retrainable software and intuitive command-line interface allow easy adaptation to other RNA classes. RISER is available at https://github.com/comprna/riser .

Generation and analysis of a mouse multi-tissue genome annotation atlas

Generating an accurate and complete genome annotation for an organism is complex because the cells within each tissue can express a unique set of transcript isoforms from a unique set of genes. A comprehensive genome annotation should contain information on what tissues express what transcript isoforms at what level. This tissue-level isoform information can then inform a wide range of research questions as well as experiment designs. Long-read sequencing technology combined with advanced full-length cDNA library preparation methods has now achieved throughput and accuracy where generating these types of annotations is achievable. Here, we show this by generating a genome annotation of the mouse (Mus musculus). We used the nanopore-based R2C2 long-read sequencing method to generate 64 million highly accurate full length cDNA consensus reads - averaging 5.4 million reads per tissue for a dozen tissues. Using the Mandalorion tool we processed these reads to generate the Tissue-level Atlas of Mouse Isoforms (TAMI - available at https://genome.ucsc.edu/s/vollmers/TAMI) which we believe will be a valuable complement to conventional, manually curated reference genome annotations.

Improving the RNA velocity approach with single-cell RNA lifecycle (nascent, mature and degrading RNAs) sequencing technologies

We presented an experimental method called FLOUR-seq, which combines BD Rhapsody and nanopore sequencing to detect the RNA lifecycle (including nascent, mature, and degrading RNAs) in cells. Additionally, we updated our HIT-scISOseq V2 to discover a more accurate RNA lifecycle using 10x Chromium and Pacbio sequencing. Most importantly, to explore how single-cell full-length RNA sequencing technologies could help improve the RNA velocity approach, we introduced a new algorithm called 'Region Velocity' to more accurately configure cellular RNA velocity. We applied this algorithm to study spermiogenesis and compared the performance of FLOUR-seq with Pacbio-based HIT-scISOseq V2. Our findings demonstrated that 'Region Velocity' is more suitable for analyzing single-cell full-length RNA data than traditional RNA velocity approaches. These novel methods could be useful for researchers looking to discover full-length RNAs in single cells and comprehensively monitor RNA lifecycle in cells.

Transcript capture and ultradeep long-read RNA sequencing (CAPLRseq) to diagnose HNPCC/Lynch syndrome

PURPOSE

Whereas most human genes encode multiple mRNA isoforms with distinct function, clinical workflows for assessing this heterogeneity are not readily available. This is a substantial shortcoming, considering that up to 25% of disease-causing gene variants are suspected of disrupting mRNA splicing or mRNA abundance. Long-read sequencing can readily portray mRNA isoform diversity, but its sensitivity is relatively low due to insufficient transcriptome penetration.

METHODS

We developed and applied capture-based target enrichment from patient RNA samples combined with Oxford Nanopore long-read sequencing for the analysis of 123 hereditary cancer transcripts (capture and ultradeep long-read RNA sequencing (CAPLRseq)).

RESULTS

Validating CAPLRseq, we confirmed 17 cases of hereditary non-polyposis colorectal cancer/Lynch syndrome based on the demonstration of splicing defects and loss of allele expression of mismatch repair genes MLH1, PMS2, MSH2 and MSH6. Using CAPLRseq, we reclassified two variants of uncertain significance in MSH6 and PMS2 as either likely pathogenic or benign.

CONCLUSION

Our data show that CAPLRseq is an automatable and adaptable workflow for effective transcriptome-based identification of disease variants in a clinical diagnostic setting.

Identifying and quantifying isoforms from accurate full-length transcriptome sequencing reads with Mandalorion

In this manuscript, we introduce and benchmark Mandalorion v4.1 for the identification and quantification of full-length transcriptome sequencing reads. It further improves upon the already strong performance of Mandalorion v3.6 used in the LRGASP consortium challenge. By processing real and simulated data, we show three main features of Mandalorion: first, Mandalorion-based isoform identification has very high precision and maintains high recall even in the absence of any genome annotation. Second, isoform read counts as quantified by Mandalorion show a high correlation with simulated read counts. Third, isoforms identified by Mandalorion closely reflect the full-length transcriptome sequencing data sets they are based on.

Sequencing Illumina libraries at high accuracy on the ONT MinION using R2C2

High-throughput short-read sequencing has taken on a central role in research and diagnostics. Hundreds of different assays take advantage of Illumina short-read sequencers, the predominant short-read sequencing technology available today. Although other short-read sequencing technologies exist, the ubiquity of Illumina sequencers in sequencing core facilities and the high capital costs of these technologies have limited their adoption. Among a new generation of sequencing technologies, Oxford Nanopore Technologies (ONT) holds a unique position because the ONT MinION, an error-prone long-read sequencer, is associated with little to no capital cost. Here we show that we can make short-read Illumina libraries compatible with the ONT MinION by using the rolling circle to concatemeric consensus (R2C2) method to circularize and amplify the short library molecules. This results in longer DNA molecules containing tandem repeats of the original short library molecules. This longer DNA is ideally suited for the ONT MinION, and after sequencing, the tandem repeats in the resulting raw reads can be converted into high-accuracy consensus reads with similar error rates to that of the Illumina MiSeq. We highlight this capability by producing and benchmarking RNA-seq, ChIP-seq, and regular and target-enriched Tn5 libraries. We also explore the use of this approach for rapid evaluation of sequencing library metrics by implementing a real-time analysis workflow.

Integrated ONT Full-Length Transcriptome and Metabolism Reveal the Mechanism Affecting Ovulation in Muscovy Duck (Cairina moschata)

Ovulation is a complicated physiological process that is regulated by a multitude of different pathways. In comparison to mammalian studies, there are few reports of ovulation in Muscovy ducks, and the molecular mechanism of ovarian development remained unclear. In order to identify candidate genes and metabolites related to Muscovy duck follicular ovulation, the study combined Oxford Nanopore Technologies (ONT) full-length transcriptome and metabolomics to analyze the differences in gene expression and metabolite accumulation in the ovaries between pre-ovulation (PO) and consecutive ovulation (CO) Muscovy ducks. 83 differentially accumulated metabolites (DAMs) were identified using metabolomics analysis, 33 of which are related to lipids. Combined with data from previous transcriptomic analyses found that DEGs and DAMs were particularly enriched in processes including the regulation of glycerophospholipid metabolism pathway, arachidonic acid metabolic pathway and the steroid biosynthetic pathway. In summary, the novel potential mechanisms that affect ovulation in Muscovy ducks may be related to lipid metabolism, and the findings provide new insights into the mechanisms of ovulation in waterfowl and will contribute to a better understanding of changes in the waterfowl ovarian development regulatory network.

Glacial ice supports a distinct and undocumented polar bear subpopulation persisting in late 21st-century sea-ice conditions

Polar bears are susceptible to climate warming because of their dependence on sea ice, which is declining rapidly. We present the first evidence for a genetically distinct and functionally isolated group of polar bears in Southeast Greenland. These bears occupy sea-ice conditions resembling those projected for the High Arctic in the late 21st century, with an annual ice-free period that is >100 days longer than the estimated fasting threshold for the species. Whereas polar bears in most of the Arctic depend on annual sea ice to catch seals, Southeast Greenland bears have a year-round hunting platform in the form of freshwater glacial mélange. This suggests that marine-terminating glaciers, although of limited availability, may serve as previously unrecognized climate refugia. Conservation of Southeast Greenland polar bears, which meet criteria for recognition as the world's 20th polar bear subpopulation, is necessary to preserve the genetic diversity and evolutionary potential of the species.

Long-read isoform sequencing reveals tissue-specific isoform expression between active and hibernating brown bears (Ursus arctos)

Understanding hibernation in brown bears (Ursus arctos) can provide insight into some human diseases. During hibernation, brown bears experience periods of insulin resistance, physical inactivity, extreme bradycardia, obesity, and the absence of urine production. These states closely mimic aspects of human diseases such as type 2 diabetes, muscle atrophy, as well as renal and heart failure. The reversibility of these states from hibernation to active season enables the identification of mediators with possible therapeutic value for humans. Recent studies have identified genes and pathways that are differentially expressed between active and hibernation seasons in bears. However, little is known about the role of differential expression of gene isoforms on hibernation physiology. To identify both distinct and novel mRNA isoforms, full-length RNA-sequencing (Iso-Seq) was performed on adipose, skeletal muscle, and liver from three individual bears sampled during both active and hibernation seasons. The existing reference genome annotation was improved by combining it with the Iso-Seq data. Short-read RNA-sequencing data from six individuals were mapped to the new reference annotation to quantify differential isoform usage (DIU) between tissues and seasons. We identified differentially expressed isoforms in all three tissues, to varying degrees. Adipose had a high level of DIU with isoform switching, regardless of whether the genes were differentially expressed. Our analyses revealed that DIU, even in the absence of differential gene expression, is an important mechanism for modulating genes during hibernation. These findings demonstrate the value of isoform expression studies and will serve as the basis for deeper exploration into hibernation biology.

Single-cell isoform analysis in human immune cells

High-throughput single-cell analysis today is facilitated by protocols like the 10X Genomics platform or Drop-Seq which generate cDNA pools in which the origin of a transcript is encoded at its 5′ or 3′ end. Here, we used R2C2 to sequence and demultiplex 12 million full-length cDNA molecules generated by the 10X Genomics platform from ~3000 peripheral blood mononuclear cells. We use these reads, independent from Illumina data, to identify B cell, T cell, and monocyte clusters and generate isoform-level transcriptomes for cells and cell types. Finally, we extract paired adaptive immune receptor sequences unique to each T and B cell.

Large-Scale Multiplexing Permits Full-Length Transcriptome Annotation of 32 Bovine Tissues From a Single Nanopore Flow Cell

A comprehensive annotation of transcript isoforms in domesticated species is lacking. Especially considering that transcriptome complexity and splicing patterns are not well-conserved between species, this presents a substantial obstacle to genomic selection programs that seek to improve production, disease resistance, and reproduction. Recent advances in long-read sequencing technology have made it possible to directly extrapolate the structure of full-length transcripts without the need for transcript reconstruction. In this study, we demonstrate the power of long-read sequencing for transcriptome annotation by coupling Oxford Nanopore Technology (ONT) with large-scale multiplexing of 93 samples, comprising 32 tissues collected from adult male and female Hereford cattle. More than 30 million uniquely mapping full-length reads were obtained from a single ONT flow cell, and used to identify and characterize the expression dynamics of 99,044 transcript isoforms at 31,824 loci. Of these predicted transcripts, 21% exactly matched a reference transcript, and 61% were novel isoforms of reference genes, substantially increasing the ratio of transcript variants per gene, and suggesting that the complexity of the bovine transcriptome is comparable to that in humans. Over 7,000 transcript isoforms were extremely tissue-specific, and 61% of these were attributed to testis, which exhibited the most complex transcriptome of all interrogated tissues. Despite profiling over 30 tissues, transcription was only detected at about 60% of reference loci. Consequently, additional studies will be necessary to continue characterizing the bovine transcriptome in additional cell types, developmental stages, and physiological conditions. However, by here demonstrating the power of ONT sequencing coupled with large-scale multiplexing, the task of exhaustively annotating the bovine transcriptome - or any mammalian transcriptome - appears significantly more feasible.

Generation of an isoform-level transcriptome atlas of macrophage activation

RNA-seq is routinely used to measure gene expression changes in response to cell perturbation. Genes upregulated or downregulated following some perturbation are designated as genes of interest, and their most expressed isoform(s) would then be selected for follow-up experimentation. However, because of its need to fragment RNA molecules, RNA-seq is limited in its ability to capture gene isoforms and their expression patterns. This lack of isoform-specific data means that isoforms would be selected based on annotation databases that are incomplete, not tissue specific, or do not provide key information on expression levels. As a result, minority or nonexistent isoforms might be selected for follow-up, leading to loss in valuable resources and time. There is therefore a great need to comprehensively identify gene isoforms along with their corresponding levels of expression. Using the long-read nanopore-based R2C2 method, which does not fragment RNA molecules, we generated an Isoform-level transcriptome Atlas of Macrophage Activation that identifies full-length isoforms in primary human monocyte-derived macrophages. Macrophages are critical innate immune cells important for recognizing pathogens through binding of pathogen-associated molecular patterns to toll-like receptors, culminating in the initiation of host defense pathways. We characterized isoforms for most moderately-to-highly expressed genes in resting and toll-like receptor–activated monocyte-derived macrophages, identified isoforms differentially expressed between conditions, and validated these isoforms by RT-qPCR. We compiled these data into a user-friendly data portal within the UCSC Genome Browser (https://genome.ucsc.edu/s/vollmers/IAMA). Our atlas represents a valuable resource for innate immune research, providing unprecedented isoform information for primary human macrophages.

Species and population specific gene expression in blood transcriptomes of marine turtles

BACKGROUND

Transcriptomic data has demonstrated utility to advance the study of physiological diversity and organisms' responses to environmental stressors. However, a lack of genomic resources and challenges associated with collecting high-quality RNA can limit its application for many wild populations. Minimally invasive blood sampling combined with de novo transcriptomic approaches has great potential to alleviate these barriers. Here, we advance these goals for marine turtles by generating high quality de novo blood transcriptome assemblies to characterize functional diversity and compare global transcriptional profiles between tissues, species, and foraging aggregations.

RESULTS

We generated high quality blood transcriptome assemblies for hawksbill (Eretmochelys imbricata), loggerhead (Caretta caretta), green (Chelonia mydas), and leatherback (Dermochelys coriacea) turtles. The functional diversity in assembled blood transcriptomes was comparable to those from more traditionally sampled tissues. A total of 31.3% of orthogroups identified were present in all four species, representing a core set of conserved genes expressed in blood and shared across marine turtle species. We observed strong species-specific expression of these genes, as well as distinct transcriptomic profiles between green turtle foraging aggregations that inhabit areas of greater or lesser anthropogenic disturbance.

CONCLUSIONS

Obtaining global gene expression data through non-lethal, minimally invasive sampling can greatly expand the applications of RNA-sequencing in protected long-lived species such as marine turtles. The distinct differences in gene expression signatures between species and foraging aggregations provide insight into the functional genomics underlying the diversity in this ancient vertebrate lineage. The transcriptomic resources generated here can be used in further studies examining the evolutionary ecology and anthropogenic impacts on marine turtles.

Error correction enables use of Oxford Nanopore technology for reference-free transcriptome analysis

Oxford Nanopore (ONT) is a leading long-read technology which has been revolutionizing transcriptome analysis through its capacity to sequence the majority of transcripts from end-to-end. This has greatly increased our ability to study the diversity of transcription mechanisms such as transcription initiation, termination, and alternative splicing. However, ONT still suffers from high error rates which have thus far limited its scope to reference-based analyses. When a reference is not available or is not a viable option due to reference-bias, error correction is a crucial step towards the reconstruction of the sequenced transcripts and downstream sequence analysis of transcripts. In this paper, we present a novel computational method to error correct ONT cDNA sequencing data, called isONcorrect. IsONcorrect is able to jointly use all isoforms from a gene during error correction, thereby allowing it to correct reads at low sequencing depths. We are able to obtain a median accuracy of 98.9-99.6%, demonstrating the feasibility of applying cost-effective cDNA full transcript length sequencing for reference-free transcriptome analysis.

Improved bacterial RNA-seq by Cas9-based depletion of ribosomal RNA reads

A major challenge for RNA-seq analysis of gene expression is to achieve sufficient coverage of informative nonribosomal transcripts. In eukaryotic samples, this is typically achieved by selective oligo(dT)-priming of messenger RNAs to exclude ribosomal RNA (rRNA) during cDNA synthesis. However, this strategy is not compatible with prokaryotes in which functional transcripts are generally not polyadenylated. To overcome this, we adopted DASH (depletion of abundant sequences by hybridization), initially developed for eukaryotic cells, to improve both the sensitivity and depth of bacterial RNA-seq. DASH uses the Cas9 nuclease to remove unwanted cDNA sequences prior to library amplification. We report the design, evaluation, and optimization of DASH experiments for standard bacterial short-read sequencing approaches, including software for automated guide RNA (gRNA) design for Cas9-mediated cleavage in bacterial rDNA sequences. Using these gRNA pools, we effectively removed rRNA reads (56%-86%) in RNA-seq libraries from two different model bacteria, the Gram-negative pathogen Salmonella enterica and the anaerobic gut commensal Bacteroides thetaiotaomicron DASH works robustly, even with subnanogram amounts of input RNA. Its efficiency, high sensitivity, ease of implementation, and low cost (∼$5 per sample) render DASH an attractive alternative to rRNA removal protocols, in particular for material-constrained studies where conventional ribodepletion techniques fail.

Complete characterization of the human immune cell transcriptome using accurate full-length cDNA sequencing

The human immune system relies on highly complex and diverse transcripts and the proteins they encode. These include transcripts encoding human leukocyte antigen (HLA) receptors as well as B cell and T cell receptors (BCR and TCR). Determining which alleles an individual possesses for each HLA gene (high-resolution HLA typing) is essential to establish donor–recipient compatibility in organ and bone marrow transplantations. In turn, the repertoires of millions of unique BCR and TCR transcripts in each individual carry a vast amount of health-relevant information. Both short-read RNA-seq-based HLA typing and BCR/TCR repertoire sequencing (AIRR-seq) currently rely on our incomplete knowledge of the genetic diversity at HLA and BCR/TCR loci. Here, we generated over 10,000,000 full-length cDNA sequences at a median accuracy of 97.9% using our nanopore sequencing-based Rolling Circle Amplification to Concatemeric Consensus (R2C2) protocol. We used this data set to (1) show that deep and accurate full-length cDNA sequencing can be used to provide isoform-level transcriptome analysis for more than 9000 loci, (2) generate accurate sequences of HLA alleles, and (3) extract detailed AIRR data for the analysis of the adaptive immune system. The HLA and AIRR analysis approaches we introduce here are untargeted and therefore do not require prior knowledge of the composition or genetic diversity of HLA and BCR/TCR loci.