Skip to main content

Unraveling changes in cellular gene expression during viral infection

Periodic Reporting for period 4 - Profile Infection (Unraveling changes in cellular gene expression during viral infection)

Reporting period: 2019-12-01 to 2020-05-31

The herpesvirus human cytomegalovirus (HCMV) infects the majority of the world's population, leading to severe diseases in millions of newborns and immunocompromised adults annually. During infection, HCMV extensively manipulates cellular gene expression to maintain conditions favorable for efficient viral propagation. Identifying the pathways that the virus relies on or subverts is of great interest as they have the potential to provide new therapeutic windows and reveal novel principles in cell biology. Over the past years high-throughput analyses have profoundly broadened our understanding of the processes that occur during HCMV infection. However, much of this analysis was focused on transcriptional changes at the lytic phase of infection leaving post-transcriptional regulation. Furthermore HCMV , like all other herpesviruses establishes latent infection which allows a lifelong infection. Although reactivation from latency causes serious disease in immunocompromised individuals, our molecular understanding of latency and our ability to capture latent cells is very limited. In our work we have implemented novel emerging technologies to extend our knowledge in to these areas that were heretofore unattainable.
This allowed us to uncover multiple novel layers of cellular gene regulation that HCMV is utilizing to modulate the host cell. These findings enhance our understanding of this sophisticated pathogen, reveal general principles in cell biology and can help to facilitate the development of novel anti-viral strategies. Furthermore, our work revealed HCMV latency is mostly driven by the host cell environment and we have characterized cell surface markers that allow enrichment of latent cells as well as the changes that occur in cells during latency. These findings provide a potential new context for deciphering virus-host interactions underlying HCMV lifelong persistence and can provide novel means to identify latent cells to prevent reactivation and disease.
Despite substantial variation in genome size, all viruses remain unconditionally dependent upon the cellular protein synthesis machinery to translate viral mRNAs. The tools we developed allowed us to unbiasedly explore how viruses manipulate host translation and how this regulation is embedded in the cellular circuitry. We showed that the response to HCMV infection is dominated by extensive transcriptional changes, but that there is also diverse and dynamic translational regulation. Furthermore by integrating our translation measurements with measurements of protein abundance we comprehensively identified host proteins that may be degraded during HCMV infection, implying biological importance (Tirosh et al. PLoS Pathogens, 2015). This unbiased analysis led us to reveal a novel antiviral activity of ROCK kinase during HCMV infection (Eliyahu et al., J. Virol 2019). Moreover we developed an approach that allows systematic identification of cis-regulatory elements in mRNAs by identifying local changes in mRNA secondary structure. Utilizing this strategy we uncovered dozens of regions that control host mRNA translation and stability during infection (Mizrahi et al. Mol Cell, 2018). Finally, we discovered a central role of m6A RNA modification as a negative regulator of type I interferon response, offering a potential unifying model for interpreting the plethora of phenotypes associated with m6A modification and propagation of diverse viruses (Winkler et al., Nature Immunology, 2019).

Although reactivation from latency causes serious disease in immunocompromised individuals, our molecular understanding of latency and ability to detect latent cells are still very limited. We utilized the power of single-cell RNA-seq (scRNA-seq) to unbiasedly examine the viral transcriptome in latency and made the ground-breaking observation that in latency there are mostly quantitative but not qualitative changes in viral gene expression (Shnayder et al. mBio, 2018). Furthermore, by exploiting the ability of scRNA-seq to tackle heterogeneity in infection level, we characterized specific changes that occur in cellular gene expression during latency. This analysis revealed that regardless of the differentiation state of the infected cells, HCMV latency drives cells towards an anergic-like monocyte state (Shnayder et al., eLife 2020).
Our findings highlight the central role of m6A as a negative regulator of type-I IFN response, by dictating the fast turnover of IFN mRNAs and consequently facilitating viral propagation (Winkler et al., Nature Immunology, 2019). In a broader sense, this work demonstrates how studying virus-host interaction can shed light on basic molecular processes of the host cell. These findings suggest a potential parsimonious unifying model for interpreting the plethora of phenotypes associated with depletion of m6A and propagation of a variety of viruses. To date, studies have illustrated the critical role of m6A machinery in diverse development stages, physiology and disease. However, the ultimate challenge is to link specific methylation sites to phenotypes. We deem IFNB to be a perfect candidate to address this challenge as it is a tightly regulated cytokine and subtle destabilization of its mRNA, caused by m6A, can potentially lead to dramatic phenotypes. To address this challenge in an ongoing work we are generating mouse and human cells in which the putative m6A-modified adenosines in IFNB are mutated. We anticipate these cells and mice may provide, for the first time, a direct connection between m6A modification on a specific transcript and a physiological phenotype.


With regard to full characterization of the latency state, one of the obstacles in studying HCMV latency is the heterogeneity in the cell culture models that are being used. We utilized the power of single-cell RNA-seq (scRNA-seq) to unbiasedly examine the viral transcriptome in latency and made the ground-breaking observation that in latency there are mostly quantitative but not qualitative changes in viral gene expression (Shnayder et al. mBio, 2018). This analysis emphasizes undeniable surprising similarity between latent and late lytic transcriptional programs. These results, together with additional analysis we conducted in natural samples, challenge the view of a well-defined HCMV latency-specific transcriptional program that is composed of few selective functional genes. Instead, these measurements raise the possibility of a gradual repression of viral gene expression resulting in low-level expression of a program that resembles late lytic infection stage.
a central withstanding question is- what are the molecular events leading to the viral transcriptional state during latency? We are currently conducting high temporal resolution measurements, at single cell level, of the early steps during latent and infection of different myeloid populations in order to shed light on the molecular events that lead to the wide but repressed viral gene expression in latent cells.