Herpes simplex virus 1 (HSV-1) is an important human pathogen and a paradigm for virus-induced host shut-off. Here we show that global changes in transcription and RNA processing and their impact on translation can be analysed in a single experimental setting by applying 4sU-tagging of newly transcribed RNA and ribosome profiling to lytic HSV-1 infection. Unexpectedly, we find that HSV-1 triggers the disruption of transcription termination of cellular, but not viral, genes. This results in extensive transcription for tens of thousands of nucleotides beyond poly(A) sites and into downstream genes, leading to novel intergenic splicing between exons of neighbouring cellular genes. As a consequence, hundreds of cellular genes seem to be transcriptionally induced but are not translated. In contrast to previous reports, we show that HSV-1 does not inhibit co-transcriptional splicing. Our approach thus substantially advances our understanding of HSV-1 biology and establishes HSV-1 as a model system for studying transcription termination.

Lytic herpes simplex virus 1 (HSV-1) infection triggers disruption of transcription termination (DoTT) of most cellular genes, resulting in extensive intergenic transcription. Similarly, cellular stress responses lead to gene-specific transcription downstream of genes (DoG). In this study, we performed a detailed comparison of DoTT/DoG transcription between HSV-1 infection, salt and heat stress in primary human fibroblasts using 4sU-seq and ATAC-seq. Although DoTT at late times of HSV-1 infection was substantially more prominent than DoG transcription in salt and heat stress, poly(A) read-through due to DoTT/DoG transcription and affected genes were significantly correlated between all three conditions, in particular at earlier times of infection. We speculate that HSV-1 either directly usurps a cellular stress response or disrupts the transcription termination machinery in other ways but with similar consequences. In contrast to previous reports, we found that inhibition of Ca2+ signaling by BAPTA-AM did not specifically inhibit DoG transcription but globally impaired transcription. Most importantly, HSV-1-induced DoTT, but not stress-induced DoG transcription, was accompanied by a strong increase in open chromatin downstream of the affected poly(A) sites. In its extent and kinetics, downstream open chromatin essentially matched the poly(A) read-through transcription. We show that this does not cause but rather requires DoTT as well as high levels of transcription into the genomic regions downstream of genes. This raises intriguing new questions regarding the role of histone repositioning in the wake of RNA Polymerase II passage downstream of impaired poly(A) site recognition.

Ribosome profiling has predicted thousands of short open reading frames (sORFs) in eukaryotic cells, but still suffers from substantial levels of noise. PRICE (https://github.com/erhard-lab/price) is a computational method modeling the experimental noise to accurately resolve overlapping sORFs and non-canonical translation initiation. We experimentally validated translation using MHC-I peptidomics and saw that sORF-derived peptides efficiently enter the MHC-I presentation pathway and thus constitute a substantial fraction of the antigen repertoire.

RNA synthesis and decay rates determine the steady-state levels of cellular RNAs. Metabolic tagging of newly transcribed RNA by 4-thiouridine (4sU) can reveal the relative contributions of RNA synthesis and decay rates. The kinetics of RNA processing, however, had so far remained unresolved. Here, we show that ultrashort 4sU-tagging not only provides snapshot pictures of eukaryotic gene expression but, when combined with progressive 4sU-tagging and RNA-seq, reveals global RNA processing kinetics at nucleotide resolution. Using this method, we identified classes of rapidly and slowly spliced/degraded introns. Interestingly, each class of splicing kinetics was characterized by a distinct association with intron length, gene length, and splice site strength. For a large group of introns, we also observed long lasting retention in the primary transcript, but efficient secondary splicing or degradation at later time points. Finally, we show that processing of most, but not all small nucleolar (sno)RNA-containing introns is remarkably inefficient with the majority of introns being spliced and degraded rather than processed into mature snoRNAs. In summary, our study yields unparalleled insights into the kinetics of RNA processing and provides the tools to study molecular mechanisms of RNA processing and their contribution to the regulation of gene expression.