Landscape of transcription in human cells

Sarah Djebali, Carrie Davis, Angelika Merkel, Alexander Dobin, Timo Lassmann, A Mortazavi, Andrea Tanzer, Julien Lagarde, Wei Lin, Felix Schlesinger, Chenghai Xue, Georgi K. Marinov, Jainab Khatun, Brian A. Williams, Chris Zaleski, Joel Rozowsky, Maik Röder, Felix Kokocinski, Rehab Abdelhamid, Tyler Alioto, Igor Antoshechkin, Michael T. Baer, Nadav Bar, Philippe Batut, Kimberly Bell, Ian Bell, Sudipto K. Chakrabortty, Xian Chen, Jacqueline Chrast, João Curado, Thomas Derrien, Jörg Drenkow, Erica Dumais, Jacqueline Dumais, Radha Duttagupta, Emilie Falconnet, Meagan Fastuca, Kata Fejes-Toth, Pedro G. Ferreira, Sylvain Foissac, Melissa J. Fullwood, Hui Gao, David González, Assaf Gordon, Harsha P. Gunawardena, Cédric Howald, Sonali Jha, Rory Johnson, Philipp Kapranov, Brandon King, Colin Kingswood, Oscar Junhong Luo, Eddie Park, Kimberly Persaud, Jonathan Preall, Paolo Ribeca, Brian A. Risk, Daniel Robyr, Michael Sammeth, Lorian Schaffer, Lei-Hoon See, Atif Shahab, Jørgen Skancke, Ana Maria Suzuki, Hazuki Takahashi, Hagen Tilgner, Diane Trout, Nathalie Walters, Huaien Wang, John A. Wrobel, Yanbao Yu, Xiaoan Ruan, Yoshihide Hayashizaki, Jennifer Harrow, Mark Gerstein, Tim Hubbard, Alexandre Reymond, Stylianos E. Antonarakis, Gregory J. Hannon, Morgan C. Giddings, Yijun Ruan, B Wold, Piero Carninci, Roderic Guigó, T GingerasNature 2012 · DOI

Abstract

Eukaryotic cells make many types of primary and processed RNAs that are found either in specific subcellular compartments or throughout the cells. A complete catalogue of these RNAs is not yet available and their characteristic subcellular localizations are also poorly understood. Because RNA represents the direct output of the genetic information encoded by genomes and a significant proportion of a cell's regulatory capabilities are focused on its synthesis, processing, transport, modification and translation, the generation of such a catalogue is crucial for understanding genome function. Here we report evidence that three-quarters of the human genome is capable of being transcribed, as well as observations about the range and levels of expression, localization, processing fates, regulatory regions and modifications of almost all currently annotated and thousands of previously unannotated RNAs. These observations, taken together, prompt a redefinition of the concept of a gene.

Introduction

For decades, molecular biology has prioritized the approximately 1–2% of the human genome that codes for proteins. The remaining "dark matter" was frequently dismissed as non-functional "junk DNA." While early pilot studies suggested transcription might be more widespread, a comprehensive, genome-wide catalog of all RNA types—including their subcellular localizations and processing fates—remained elusive. This study, part of the ENCODE project, seeks to characterize the full landscape of transcription to understand how the genome functions as a whole.

Methodology

The researchers employed a multi-pronged experimental approach across 15 diverse human cell lines (including K562, GM12878, and HeLa-S3). Key techniques included:

  • Subcellular Fractionation: Separating cellular components into cytosol and nucleus to map RNA localization.
  • Transcriptome Sequencing (RNA-seq): Sequencing long RNAs (>200 nucleotides), categorized into polyadenylated (PolyA+) and non-polyadenylated (PolyA-) fractions.
  • Termini Mapping: Utilizing CAGE (Cap Analysis of Gene Expression) for 5’ start sites and PET (Paired-End Tags) for both 5’ and 3’ ends.
  • De Novo Assembly: Applying the Cufflinks algorithm to assemble novel transcripts from raw sequence data.
  • Validation: Using mass spectrometry to detect protein products and targeted sequencing to verify predicted splice sites.

Key Results

The study reveals a genome that is nearly continuously active. Approximately 74.7% of the genome is covered by primary transcripts, while 62.1% is covered by processed transcripts. This pervasive transcription significantly shrinks the perceived "intergenic" space, as thousands of novel RNAs overlap or interleave with known genes.

Proportion of RNA features detected across subcellular fractions and polyadenylation states.

As shown in the figure above, the detection of features such as splice junctions, exons, and transcripts varies significantly across subcellular compartments. Protein-coding transcripts are found predominantly in the cytosol, whereas long non-coding RNAs (lncRNAs) are largely sequestered in the nucleus. Furthermore, the study found that genes typically express multiple isoforms simultaneously, often plateauing at 10–12 versions per gene per cell line. Splicing was observed to occur almost immediately on the chromatin during synthesis.

Discussion and Redefining the Gene

The discovery of thousands of novel transcripts—increasing the number of annotated genes by 80%—challenges the traditional definition of a gene as a discrete, standalone unit of heredity. Many of these new additions are non-coding RNAs, including eRNAs transcribed from enhancers, which are primarily nuclear and cell-type specific.

While the study proves that three-quarters of the genome is transcribed, the functional significance of every transcript remains a subject for future research. However, the high degree of organization, specific localization, and complex processing suggests that much of this activity is functional rather than mere "transcriptional noise."

Conclusion

The ENCODE project's mapping of the human transcriptome reveals a complex, overlapping, and pervasive landscape of transcription. By documenting the localization and modification of both known and novel RNAs, this work provides a crucial foundation for understanding the regulatory capabilities of the human genome and necessitates a conceptual shift in how we define genetic units.

Summary updated

Abstract

Eukaryotic cells make many types of primary and processed RNAs that are found either in specific subcellular compartments or throughout the cells. A complete catalogue of these RNAs is not yet available and their characteristic subcellular localizations are also poorly understood. Because RNA represents the direct output of the genetic information encoded by genomes and a significant proportion of a cell's regulatory capabilities are focused on its synthesis, processing, transport, modification and translation, the generation of such a catalogue is crucial for understanding genome function. Here we report evidence that three-quarters of the human genome is capable of being transcribed, as well as observations about the range and levels of expression, localization, processing fates, regulatory regions and modifications of almost all currently annotated and thousands of previously unannotated RNAs. These observations, taken together, prompt a redefinition of the concept of a gene.
BibTeX
@inproceedings{djebali2012landscape,
  author = {Sarah Djebali and Carrie Davis and Angelika Merkel and Alexander Dobin and Timo Lassmann and A Mortazavi and Andrea Tanzer and Julien Lagarde and Wei Lin and Felix Schlesinger and Chenghai Xue and Georgi K. Marinov and Jainab Khatun and Brian A. Williams and Chris Zaleski and Joel Rozowsky and Maik Röder and Felix Kokocinski and Rehab Abdelhamid and Tyler Alioto and Igor Antoshechkin and Michael T. Baer and Nadav Bar and Philippe Batut and Kimberly Bell and Ian Bell and Sudipto K. Chakrabortty and Xian Chen and Jacqueline Chrast and João Curado and Thomas Derrien and Jörg Drenkow and Erica Dumais and Jacqueline Dumais and Radha Duttagupta and Emilie Falconnet and Meagan Fastuca and Kata Fejes-Toth and Pedro G. Ferreira and Sylvain Foissac and Melissa J. Fullwood and Hui Gao and David González and Assaf Gordon and Harsha P. Gunawardena and Cédric Howald and Sonali Jha and Rory Johnson and Philipp Kapranov and Brandon King and Colin Kingswood and Oscar Junhong Luo and Eddie Park and Kimberly Persaud and Jonathan Preall and Paolo Ribeca and Brian A. Risk and Daniel Robyr and Michael Sammeth and Lorian Schaffer and Lei-Hoon See and Atif Shahab and Jørgen Skancke and Ana Maria Suzuki and Hazuki Takahashi and Hagen Tilgner and Diane Trout and Nathalie Walters and Huaien Wang and John A. Wrobel and Yanbao Yu and Xiaoan Ruan and Yoshihide Hayashizaki and Jennifer Harrow and Mark Gerstein and Tim Hubbard and Alexandre Reymond and Stylianos E. Antonarakis and Gregory J. Hannon and Morgan C. Giddings and Yijun Ruan and B Wold and Piero Carninci and Roderic Guigó and T Gingeras},
  title = {{Landscape of transcription in human cells}},
  booktitle = {Nature},
  year = {2012},
  doi = {10.1038/nature11233},
}
Figure