Spatially resolved clonal copy number alterations in benign and malignant tissue.

Defining the transition from benign to malignant tissue is fundamental to improving early diagnosis of cancer1. Here we use a systematic approach to study spatial genome integrity in situ and describe previously unidentified clonal relationships. We used spatially resolved transcriptomics2 to infer spatial copy number variations in >120,000 regions across multiple organs, in benign and malignant tissues. We demonstrate that genome-wide copy number variation reveals distinct clonal patterns within tumours and in nearby benign tissue using an organ-wide approach focused on the prostate. Our results suggest a model for how genomic instability arises in histologically benign tissue that may represent early events in cancer evolution. We highlight the power of capturing the molecular and spatial continuums in a tissue context and challenge the rationale for treatment paradigms, including focal therapy.

Super-resolved spatial transcriptomics by deep data fusion.

Current methods for spatial transcriptomics are limited by low spatial resolution. Here we introduce a method that integrates spatial gene expression data with histological image data from the same tissue section to infer higher-resolution expression maps. Using a deep generative model, our method characterizes the transcriptome of micrometer-scale anatomical features and can predict spatial gene expression from histology images alone.

Integrating spatial gene expression and breast tumour morphology via deep learning.

Spatial transcriptomics allows for the measurement of RNA abundance at a high spatial resolution, making it possible to systematically link the morphology of cellular neighbourhoods and spatially localized gene expression. Here, we report the development of a deep learning algorithm for the prediction of local gene expression from haematoxylin-and-eosin-stained histopathology images using a new dataset of 30,612 spatially resolved gene expression data matched to histopathology images from 23 patients with breast cancer. We identified over 100 genes, including known breast cancer biomarkers of intratumoral heterogeneity and the co-localization of tumour growth and immune activation, the expression of which can be predicted from the histopathology images at a resolution of 100 µm. We also show that the algorithm generalizes well to The Cancer Genome Atlas and to other breast cancer gene expression datasets without the need for re-training. Predicting the spatially resolved transcriptome of a tissue directly from tissue images may enable image-based screening for molecular biomarkers with spatial variation.

Spatially Resolved Transcriptomics Enables Dissection of Genetic Heterogeneity in Stage III Cutaneous Malignant Melanoma.

Cutaneous malignant melanoma (melanoma) is characterized by a high mutational load, extensive intertumoral and intratumoral genetic heterogeneity, and complex tumor microenvironment (TME) interactions. Further insights into the mechanisms underlying melanoma are crucial for understanding tumor progression and responses to treatment. Here we adapted the technology of spatial transcriptomics (ST) to melanoma lymph node biopsies and successfully sequenced the transcriptomes of over 2,200 tissue domains. Deconvolution combined with traditional approaches for dimensional reduction of transcriptome-wide data enabled us to both visualize the transcriptional landscape within the tissue and identify gene expression profiles linked to specific histologic entities. Our unsupervised analysis revealed a complex spatial intratumoral composition of melanoma metastases that was not evident through morphologic annotation. Each biopsy showed distinct gene expression profiles and included examples of the coexistence of multiple melanoma signatures within a single tumor region as well as shared profiles for lymphoid tissue characterized according to their spatial location and gene expression profiles. The lymphoid area in close proximity to the tumor region displayed a specific expression pattern, which may reflect the TME, a key component to fully understanding tumor progression. In conclusion, using the ST technology to generate gene expression profiles reveals a detailed landscape of melanoma metastases. This should inspire researchers to integrate spatial information into analyses aiming to identify the factors underlying tumor progression and therapy outcome.Significance: Applying ST technology to gene expression profiling in melanoma lymph node metastases reveals a complex transcriptional landscape in a spatial context, which is essential for understanding the multiple components of tumor progression and therapy outcome. Cancer Res; 78(20); 5970-9. ©2018 AACR.

Spatial maps of prostate cancer transcriptomes reveal an unexplored landscape of heterogeneity.

Intra-tumor heterogeneity is one of the biggest challenges in cancer treatment today. Here we investigate tissue-wide gene expression heterogeneity throughout a multifocal prostate cancer using the spatial transcriptomics (ST) technology. Utilizing a novel approach for deconvolution, we analyze the transcriptomes of nearly 6750 tissue regions and extract distinct expression profiles for the different tissue components, such as stroma, normal and PIN glands, immune cells and cancer. We distinguish healthy and diseased areas and thereby provide insight into gene expression changes during the progression of prostate cancer. Compared to pathologist annotations, we delineate the extent of cancer foci more accurately, interestingly without link to histological changes. We identify gene expression gradients in stroma adjacent to tumor regions that allow for re-stratification of the tumor microenvironment. The establishment of these profiles is the first step towards an unbiased view of prostate cancer and can serve as a dictionary for future studies.

Spatial detection of fetal marker genes expressed at low level in adult human heart tissue.

Heart failure is a major health problem linked to poor quality of life and high mortality rates. Hence, novel biomarkers, such as fetal marker genes with low expression levels, could potentially differentiate disease states in order to improve therapy. In many studies on heart failure, cardiac biopsies have been analyzed as uniform pieces of tissue with bulk techniques, but this homogenization approach can mask medically relevant phenotypes occurring only in isolated parts of the tissue. This study examines such spatial variations within and between regions of cardiac biopsies. In contrast to standard RNA sequencing, this approach provides a spatially resolved transcriptome- and tissue-wide perspective of the adult human heart, and enables detection of fetal marker genes expressed by minor subpopulations of cells within the tissue. Analysis of patients with heart failure, with preserved ejection fraction, demonstrated spatially divergent expression of fetal genes in cardiac biopsies.

Binding site discovery from nucleic acid sequences by discriminative learning of hidden Markov models.

We present a discriminative learning method for pattern discovery of binding sites in nucleic acid sequences based on hidden Markov models. Sets of positive and negative example sequences are mined for sequence motifs whose occurrence frequency varies between the sets. The method offers several objective functions, but we concentrate on mutual information of condition and motif occurrence. We perform a systematic comparison of our method and numerous published motif-finding tools. Our method achieves the highest motif discovery performance, while being faster than most published methods. We present case studies of data from various technologies, including ChIP-Seq, RIP-Chip and PAR-CLIP, of embryonic stem cell transcription factors and of RNA-binding proteins, demonstrating practicality and utility of the method. For the alternative splicing factor RBM10, our analysis finds motifs known to be splicing-relevant. The motif discovery method is implemented in the free software package Discrover. It is applicable to genome- and transcriptome-scale data, makes use of available repeat experiments and aside from binary contrasts also more complex data configurations can be utilized.

Integrative analysis revealed the molecular mechanism underlying RBM10-mediated splicing regulation.

RBM10 encodes an RNA binding protein. Mutations in RBM10 are known to cause multiple congenital anomaly syndrome in male humans, the TARP syndrome. However, the molecular function of RBM10 is unknown. Here we used PAR-CLIP to identify thousands of binding sites of RBM10 and observed significant RBM10-RNA interactions in the vicinity of splice sites. Computational analyses of binding sites as well as loss-of-function and gain-of-function experiments provided evidence for the function of RBM10 in regulating exon skipping and suggested an underlying mechanistic model, which could be subsequently validated by minigene experiments. Furthermore, we demonstrated the splicing defects in a patient carrying an RBM10 mutation, which could be explained by disrupted function of RBM10 in splicing regulation. Overall, our study established RBM10 as an important regulator of alternative splicing, presented a mechanistic model for RBM10-mediated splicing regulation and provided a molecular link to understanding a human congenital disorder.

doRiNA: a database of RNA interactions in post-transcriptional regulation.

In animals, RNA binding proteins (RBPs) and microRNAs (miRNAs) post-transcriptionally regulate the expression of virtually all genes by binding to RNA. Recent advances in experimental and computational methods facilitate transcriptome-wide mapping of these interactions. It is thought that the combinatorial action of RBPs and miRNAs on target mRNAs form a post-transcriptional regulatory code. We provide a database that supports the quest for deciphering this regulatory code. Within doRiNA, we are systematically curating, storing and integrating binding site data for RBPs and miRNAs. Users are free to take a target (mRNA) or regulator (RBP and/or miRNA) centric view on the data. We have implemented a database framework with short query response times for complex searches (e.g. asking for all targets of a particular combination of regulators). All search results can be browsed, inspected and analyzed in conjunction with a huge selection of other genome-wide data, because our database is directly linked to a local copy of the UCSC genome browser. At the time of writing, doRiNA encompasses RBP data for the human, mouse and worm genomes. For computational miRNA target site predictions, we provide an update of PicTar predictions.

The SNF2-like helicase HELLS mediates E2F3-dependent transcription and cellular transformation.

The activating E2F-transcription factors are best known for their dependence on the Retinoblastoma protein and their role in cellular proliferation. E2F3 is uniquely amplified in specific human tumours where its expression is inversely correlated with the survival of patients. Here, E2F3B interaction partners were identified by mass spectrometric analysis. We show that the SNF2-like helicase HELLS interacts with E2F3A in vivo and cooperates with its oncogenic functions. Depletion of HELLS severely perturbs the induction of E2F-target genes, hinders cell-cycle re-entry and growth. Using chromatin immmunoprecipitation coupled to sequencing, we identified genome-wide targets of HELLS and E2F3A/B. HELLS binds promoters of active genes, including the trithorax-related MLL1, and co-regulates E2F3-dependent genes. Strikingly, just as E2F3, HELLS is overexpressed in human tumours including prostate cancer, indicating that either factor may contribute to the malignant progression of tumours. Our work reveals that HELLS is important for E2F3 in tumour cell proliferation.

Select microRNAs are essential for early development in the sea urchin.

microRNAs (miRNAs) are small noncoding RNAs that mediate post-transcriptional gene regulation and have emerged as essential regulators of many developmental events. The transcriptional network during early embryogenesis of the purple sea urchin, Strongylocentrotus purpuratus, is well described and can serve as an excellent model to test functional contributions of miRNAs in embryogenesis. We examined the loss of function phenotypes of major components of the miRNA biogenesis pathway. Inhibition of de novo synthesis of Drosha and Dicer in the embryo led to consistent developmental defects, a failure to gastrulate, and embryonic lethality, including changes in the steady state levels of transcription factors and signaling molecules involved in germ layer specification. We annotated and profiled small RNA expression from the ovary and several early embryonic stages by deep sequencing followed by computational analysis. miRNAs as well as a large population of putative piRNAs (piwi-interacting RNAs) had dynamic accumulation profiles through early development. Defects in morphogenesis caused by loss of Drosha could be rescued with four miRNAs. Taken together our results indicate that post-transcriptional gene regulation directed by miRNAs is functionally important for early embryogenesis and is an integral part of the early embryonic gene regulatory network in S. purpuratus.

Large-scale sorting of C. elegans embryos reveals the dynamics of small RNA expression.

Caenorhabditis elegans is one of the most prominent model systems for embryogenesis, but collecting many precisely staged embryos has been impractical. Thus, early C. elegans embryogenesis has not been amenable to most high-throughput genomics or biochemistry assays. To overcome this problem, we devised a method to collect staged C. elegans embryos by fluorescence-activated cell sorting (eFACS). In a proof-of-principle experiment, we found that a single eFACS run routinely yielded tens of thousands of almost perfectly staged 1-cell stage embryos. As the earliest embryonic events are driven by posttranscriptional regulation, we combined eFACS with second-generation sequencing to profile the embryonic expression of small, noncoding RNAs. We discovered complex and orchestrated changes in the expression between and within almost all classes of small RNAs, including microRNAs and 26G-RNAs, during embryogenesis.

Reexamining microRNA site accessibility in Drosophila: a population genomics study.

Kertesz et al. (Nature Genetics 2008) described PITA, a miRNA target prediction algorithm based on hybridization energy and site accessibility. In this note, we used a population genomics approach to reexamine their data and found that the PITA algorithm had lower specificity than methods based on evolutionary conservation at comparable levels of sensitivity.We also showed that deeply conserved miRNAs tend to have stronger hybridization energies to their targets than do other miRNAs. Although PITA had higher specificity in predicting targets than a naïve seed-match method, this signal was primarily due to the use of a single cutoff score for all miRNAs and to the observed correlation between conservation and hybridization energy. Overall, our results clarify the accuracy of different miRNA target prediction algorithms in Drosophila and the role of site accessibility in miRNA target prediction.

Discovering microRNAs from deep sequencing data using miRDeep.

The capacity of highly parallel sequencing technologies to detect small RNAs at unprecedented depth suggests their value in systematically identifying microRNAs (miRNAs). However, the identification of miRNAs from the large pool of sequenced transcripts from a single deep sequencing run remains a major challenge. Here, we present an algorithm, miRDeep, which uses a probabilistic model of miRNA biogenesis to score compatibility of the position and frequency of sequenced RNA with the secondary structure of the miRNA precursor. We demonstrate its accuracy and robustness using published Caenorhabditis elegans data and data we generated by deep sequencing human and dog RNAs. miRDeep reports altogether approximately 230 previously unannotated miRNAs, of which four novel C. elegans miRNAs are validated by northern blot analysis.