Joan and Tom Steitz RNA Fellows Program

Spatial transcriptomics has emerged as a powerful tool for dissecting cellular heterogeneities within their native tissue context but as of today is predominantly confined to mRNA examination. Yet, the life of RNA molecules is multifaceted and dynamic, requiring spatial profiling of different RNA species throughout the life cycle to delve into the intricate RNA biology in complex tissues. Formalin-fixed paraffin-embedded (FFPE) tissues are essential in clinical practice, being the backbone of all human biopsy diagnoses. The capacity to spatially explore RNA biology in FFPE tissues holds transformative potential for pathology research. Nevertheless, the RNA within these samples is susceptible to degradation during the paraffin-embedding process and may further experience heightened degradation under suboptimal storage conditions. Additionally, RNA may undergo chemical modifications, resulting in fragmentation or resistance to the enzymatic reactions required for sequencing. The loss of poly-A tails introduces another layer of complexity, restricting the use of oligo-dT primed reverse transcription.

In Dr. Rong Fan’s lab, I lead the development of Patho-DBiT, an innovative technology tailored for spatial whole transcriptome sequencing meticulously crafted to address the distinctive challenges of clinically archived FFPE tissues. Patho-DBiT integrates in situ polyadenylation, deterministic barcoding in tissue using microfluidic chips, and computational innovations to navigate and decode RNA regulations inherent in FFPE samples. Capitalizing on the inhibitory impact of formalin fixation on endogenous endonuclease activity and RNA fragmentation naturally occurring in FFPE specimens, Patho-DBiT even outperforms fresh frozen tissue spatial transcriptomics and further allows for the profiling of a broad spectrum of RNA species.

Patho-DBiT permits spatial co-profiling of gene expression and alternative splicing, unveiling region-specific isoforms in the mouse brain. High-sensitivity transcriptomics is constructed from 5-year archived T-cell lymphoma tissues, with cross-validation conducted using super-resolution spatial phenotyping technology (CODEX). Furthermore, genome-scale single nucleotide RNA variants are captured to autonomously distinguish malignant from non-malignant cells in B-cell lymphomas. Patho-DBiT also enables spatially resolved co-profiling of large and small RNAs, facilitating the analysis of a microRNA-mRNA regulatory network within clinical biopsies and elucidating their roles in tumorigenesis. With superior intronic read capture efficiency, Patho-DBiT spatially maps RNA splicing dynamics associated with the developmental trajectory of tumor cells. High- resolution Patho-DBiT with a 10-µm spot size reveals the heterogeneities of human lymphomas within a neighborhood and traces the spatiotemporal molecular kinetics driving tumor progression at the cellular level. Patho-DBiT represents a first-of-its-kind technology, enabling the spatial exploration of rich RNA biology in FFPE tissues to aid pathology diagnosis.

In a version of life derived from RNA, RNA molecules took on the catalytic or regulatory functions typically performed by proteins in modern biology. Long non-coding RNAs (ncRNAs) maintained in bacteria are believed to be remnants of this ‘RNA World,’ and as such often participate in conserved and fundamental biological processes. As a high school student, I was given the opportunity through a public-school outreach program to conduct research at the University of Pittsburgh. Over my three years studying conserved energy pathways in yeast, I fell in love with ancient biology and the RNA molecules which often lie at its core. One year, I even dressed up as an essential regulatory process, ‘alternative splicing,’ for Halloween.

In October of 2021, I joined the Breaker Lab to further explore my interests in the RNA World and molecular evolution of life. Alongside incredible mentors, I have elucidated novel roles that non-coding RNAs likely played in regulating central carbon metabolism or metal homeostasis in early life. My current project concerns one of the largest and most widespread classes of bacteria ncRNAs whose exact biochemical function remains unknown: the Ornate, Large, Extremophilic (OLE) RNA. OLE RNAs form a ribonucleoprotein (RNP) complex which is required for cells to adapt to diverse environmental stresses such non-glucose carbon sources or excessive magnesium concentrations. As such, the Breaker Lab has recently theorized that the OLE RNP complex acts as a master regulator of cell growth in bacteria.

Previous genetic screens involving the OLE RNA suggest cells without a functional OLE RNP complex experience more ribosome stalling events. This observation motivated my hypothesis that the OLE RNA controls cell growth under stress by affecting ribosome stalling. Thanks to Tom Steitz’s careful work characterizing the mechanisms of action of several ribosome-targeting antibiotics, I assessed the effect of inducing ribosome-stalling in cells lacking the OLE RNP complex. Surprisingly, I found that in the absence of OLE RNA, cells possessed increased resistance to ribosome-stalling antibiotics.

The precedent for the role of ncRNAs in gene regulation and stress response makes these macromolecules a largely unexplored avenue to antibiotic resistance in bacteria. As I pursue a Ph.D. in biochemistry, I hope to continue to study rare and ancient RNAs. Investigation into classes like OLE RNA may uncover common biochemical phenomenon which underly paths to stress adaptation and antibiotic resistance across bacterial species.

Prior to the discovery of non-coding RNAs (ncRNAs), it was assumed that non protein-coding genetic material did not possess a functional purpose. However, decades of scientific research unveiled diverse classes of ncRNAs that are crucial for numerous molecular pathways in prokaryotes, eukaryotes, and even in viruses. As viruses have a constrained genome, any viral transcriptional output likely has a function to ensure successful infection and proliferation in the host. Introduced to the intricacies of ncRNAs in my biochemistry courses, I was fascinated by ncRNAs’ therapeutic potential and the enduring mysteries on their function, which has led me to the robust RNA community at Yale where I investigate the role of viral ncRNAs in Epstein-Barr virus (EBV) in Dr. Steitz’s lab.

EBV is a double-stranded DNA tumor virus that infects B-cells with a widespread influence, having infected over 90% worldwide. As an oncogenic herpesvirus, EBV entails a lifelong infection with increased risks of developing Burkitt’s lymphoma, Hodgkin lymphoma, and nasopharyngeal carcinoma. Interestingly, EBV encodes various ncRNAs, even during latency when gene expression is minimal. Most notably, EBV encodes an abundance of ncRNA EBER2, which regulates the expression of latent genes necessary for B-cell transformation and is crucial to EBV malignancy.

As part of my research in the Steitz lab, I investigate EBER2's role in maintaining latency and inducing lytic replication, which remains largely unknown. Specifically, I seek to understand the composition and function of the ribonucleoprotein (RNP) complex consisting of EBER2, host transcription factor PAX5, and the terminal repeat (TR) region of the EBV genome. EBER2 and host transcription factor PAX5, which is a major B-cell regulator, was discovered to indirectly colocalize to the TR region. As in-vivo crosslinking and high-throughput sequencing methods identified a novel non-coding RNA transcript as the potential intermediary that recruits EBER2 and PAX5 to the TR region of the viral genome, my current work seeks to characterize this novel transcript. Recently, I have successfully identified the start sites of several alternatively spliced variants of this novel ncRNA and aim to further investigate how particular spliced variants may contribute to the RNP complex’s regulation of EBV life cycle.

The prospect of my research synergizes both scientific and societal implications. Further understanding of how ncRNAs contribute to EBV regulation and malignant growth will expand our growing knowledge of ncRNAs’ roles in viruses, while elucidating potential antiviral strategies that could alleviate EBV’s global impact and oncogenic ramifications.

Through base pairing, cellular RNAs adopt unique functional structures that determine and regulate their catalytic, structural, and coding potential. Understanding base pairing is therefore crucial to harnessing the full potential of RNA, e.g. for therapeutic purposes. As such, one of the most fundamental questions in RNA biology remains mostly open: Base pairing within cells is currently hard to reliably predict based on sequence alone.

When I started my PhD in Karla Neugebauer’s lab, I realized that nevertheless, many researchers report that the predicted stability of the most stable RNA structure from the ensemble of all possible structures (MFE structure) correlates with some experimental read-out. Translation, splicing, degradation, localization – all correlated with RNA structure. Rarely do we find molecular mechanisms behind those correlations. I hypothesize that our understanding might be limited by a prevailing model of cellular RNA that is starting to be replaced. In an RNA version of the circular reasoning fallacy, we often depict RNA as adopting one defined structure, since we tend to study the structure of RNAs with defined structural states; those are easier to solve. Most RNA species however, especially mRNAs with evolutionary pressure on protein rather than RNA structure, are more likely to co-exist in many distinct base pairing patterns with different stabilities. Therefore, what correlates with translation, localization, etc. of an RNA may not be the stability of one of its predicted structures, but rather its base pairing potential. This might often be well- approximated by the stability of one structure from the ensemble, but crucially, the concept behind the measured variables is different and might affect our interpretation.

In my current research, I try to approach this issue by studying how base pairs first arise. By developing a new method that simultaneously detects pairing status and RNA polymerase (Pol) position on individual nascent transcripts, I show that RNA can base pair directly after exit from Pol. I show that in pre-rRNA, dynamic local base pairing predominates leading to structures that differ vastly from mature rRNA, which relies on long-range interactions that cannot form co- transcriptionally. In contrast, nascent pre-mRNA appears very similar to mRNA, suggesting that mRNA structures are dominated by small-scale local interactions which might be dictated by the process of transcription itself.

In my future academic work, I would like to continue adding to conceptual models for mRNA structure in a cellular context that incorporate RNA’s dynamic, metastable nature more explicitly, hopefully allowing us to explain more of our correlations with molecular mechanisms.

Life requires spatial and temporal control of protein production. The protein output of a given transcript reflects the integration of translation and mRNA stability. My work seeks to understand how cis-regulatory elements encoded in the untranslated regions of mRNAs regulate translation. To this end, I have developed NaP-TRAP, Nascent Peptide Translating Ribosome Affinity Purification, a Massively Parallel Reporter Assay (MPRA) that measures translation through the immunocapture of the epitope tagged-nascent chain complexes of reporter mRNAs. Using this assay, I have investigated the regulatory potential of endogenous and synthetic UTRs in the developing zebrafish embryo and human cell lines.

Given that initiation is the rate limiting step of translation, my worked has focused predominately on characterizing 5’ UTRs. In canonical eukaryotic translation initiation, the small ribosomal subunit is recruited to the 5’ UTR by eukaryotic initiation factors following cap-recognition. The subunit then scans along the 5’ UTR until it encounters a start codon, resulting in the subsequent recruitment of the large ribosomal subunit and the initiation of translation elongation. Cis-regulatory elements in the 5’ UTR can affect each step of this process. Using NaP-TRAP, I have identified universal activators and repressors as well as developmentally dynamic regulators of translation in the 5’ UTR. Yet this experience has left me with more questions than answers. For example, the reporter mRNAs that I tested were fragments of endogenous 5’ UTRs, yet some of the reporters exhibited extremely low translation values. This raises an interesting question: how are endogenous transcripts with multiple of the repressive elements in their 5’ UTRs translated at all? Perhaps these mRNAs exhibit cell-specific translation activation or serve a functional role beyond the protein for which they encode.

As I think about the future, the potential of mRNA therapeutics excites me. mRNA provides a highly customizable treatment modality without the risks and expense associated with genomic editing. Further, the regulatory potential of RNA is limitless. There are more potential sequence combinations in a 150 nucleotide segment of RNA than there are particles in the known universe. This expanse presents an exciting opportunity for technological innovation. Within the relatively small portion of the sequence space that life has sampled, there exists incredibly intricate modes of regulation. Advances in MPRAs, sequencing, and machine learning will enable us to explore this massive space in a systematic and rational manner. Who knows what we will find.

My fascination with RNA structure began during college when I first performed an in silico design of RNA-like motifs that could disrupt the frame-shifting pseudoknot of SARS-CoV-2. This experience oriented me towards exploring the crucial roles RNA structures play in translation, replication, and immune evasion in viruses. Flaviviruses, which includes Dengue, Zika, and West Nile, are viruses that possesses a vital RNA structure. These RNA viruses lead to severe diseases, such as hepatitis, encephalitis, and congenital abnormalities. With over 400 million annual infections globally, there are no pan-flaviviral therapeutics to address this great global health threat. To create effective RNA targeted therapeutics, we need to fill the critical gap in our understanding of RNA structures in viral pathogenesis and how to leverage them.

In Dr. Anna Pyle’s Lab, I focus on characterizing the architecture of the Flaviviral RNA genome. I realized that the development of an effective treatment first requires a complete structure of the subgenomic-flaviviral RNA (sfRNA). These highly conserved subgenomic RNAs are necessary for innate immune evasion, virus-induced apoptosis, and neuropathogenicity. Structurally, the sfRNA contains four pseudoknots that aid in its compaction, so I systematically interrogated the secondary structure for each pseudoknot using SHAPE-MaP. I identified a specific hierarchy in pseudoknot formation and found that the disruption of one pseudoknot could detrimentally affect sfRNA formation. Using Terbium-seq to further probe the tertiary structure, I found additional motifs that were conserved across multiple Flaviviruses, suggesting that these motifs could be potential targets for a pan-flaviviral therapeutic.

Currently, I am using cryo-EM to obtain a high-resolution structure of the sfRNA. I hope to use this structure to conclusively pinpoint potential drug binding targets. Leveraging this structure, I can further elucidate sfRNA cellular mechanisms and roles in flavivirus microevolution. I believe these structural insights will inform therapeutic development and allow us to predict future outbreaks. This work will also expand the repertoire of 3D viral RNA structures and revolutionize current understanding of structures influencing the viral lifecycle. My robust pipeline could be applied to other viral RNA systems, guiding the design of novel viral therapeutics. In line with my long-term scientific goals, I aspire to explore RNA structure- function relationships in human diseases in non-viral contexts, such as how long noncoding RNA structures may be involved in cancer. Through this structure-first approach, I aim to unravel principles of RNA structure and pave the way for the next generation of RNA therapeutics.

Within the diverse mechanisms at play in the expanding population of aging individuals, disrupted RNA processing has emerged as a prevalent theme, revealing a critical aspect of pathogenesis.

My journey into this fascinating realm began during my postgraduate research in Dr. Pallavi Gopal’s lab at Yale, where I delved into the intricate world of mRNA and RNA-binding proteins (RBPs) in the context of neurodegenerative diseases. I became captivated by the complexities of ALS, characterized by the mislocalization and aggregation of TDP-43, an RBP crucial for post-transcriptional RNA processing. My research uncovered complex interactions between TDP-43 and Ataxin-2, a polyglutamine (polyQ)-containing RBP also implicated in ALS. I found that Ataxin-2 polyQ expansions aberrantly sequester TDP-43, disrupt dynamic anterograde transport of TDP-43 ribonucleoprotein (RNP) condensates along axons, and increase their propensity to transform into pathologic aggregates. I discovered that translation of mRNAs critical for axonal and cytoskeletal integrity was suppressed in neurons expressing Ataxin-2 with polyQ expansions, providing mechanistic insights into the distal axonopathy associated with neurodegeneration.

Now, as a graduate student in Dr. Junjie Guo’s lab, I continue fueling my passion for RNA research. Working with Suzhou Yang, another graduate student, we discovered the mechanism by which an intronic nucleotide repeat expansion (NRE) in the C9ORF72 gene may encode toxic dipeptide repeat proteins in ALS and frontotemporal dementia (FTD). By capturing and sequencing NRE-containing RNAs from patient-derived cells, we elucidated that the C9ORF72 NRE is exonized by the usage of downstream 5' splice sites and exported from the nucleus in a variety of aberrantly spliced mRNA isoforms. We also found that C9ORF72 aberrant splicing was substantially elevated in both in vitro differentiated motor neurons and post-mortem brain tissues. Following up on this work, I aim to identify common RNA risk factors that may unify ALS/FTD and other age-related neurodegenerative diseases, using directly differentiated neurons and organoid models that can preserve the epigenetic age of individuals. My goal is to understand the role of mRNA (mis)splicing in both convergent and divergent disease phenotypes.

Through the years my passion for RNA research has only intensified. Becoming part of the Joan and Tom Steitz RNA Fellows Program would serve as a catalyst for my future discoveries and broaden my horizon as an RNA biologist. By unraveling the molecular intricacies of RNA dysregulation, I aim to lay the groundwork for future development of targeted therapeutics and effective interventions in the relentless battle against neurodegenerative diseases.

RNA biology and glycobiology were considered domains exclusive to each other until the discovery of glycoRNA in 2021. In recent studies, RNAs, including glycoRNAs, has been identified on mammalian cell surface, a topologically distinct space from the nucleus/cytoplasm, where the majority of cellular RNAs are traditionally found. Those findings raise numerous intriguing questions: What are the functions of cell surface RNA/glycoRNA? How are those RNAs transported and located on cell membrane? What is the chemical nature of glycoRNA... Addressing these questions will undoubtedly be of paramount importance for both the science and RNA society.

Since joining the Lu lab in 2021, I used neutrophil to study the cell surface RNAs/glycoRNAs. My recent study revealed the critical role of cell surface RNAs in neutrophil recruitment to inflammatory sites. Neutrophils contain glycoRNAs, primarily small RNAs mapping to noncoding transcripts. These glycoRNAs, predominantly located on the cell surface, play a vital role in governing the initial capture and rolling of neutrophils on blood vessel walls through the recognition by endothelial P-seletin. I also found that glycoRNAs are surprisingly stable, they are produced and transported to cell surface in a cell autonomous manner. Knockdown of murine Sidt RNA transporters genes, abolishes neutrophil glycoRNAs and functionally mimics the loss of cell surface RNAs. In summary, our study illustrates the biological importance of cell surface glycoRNAs, shedding light on the regulatory pathway governing glycoRNA production.

As the cell surface RNA/glycoRNA field is at nascent stage, many exciting questions require future explorations to address. My focus will center on several key aspects, including deciphering the code for RNA glycosylation, identifying specific types of glycoRNA that interact with P-selectin, understanding the chemical nature of glycoRNAs, developing methods for labeling and imaging cell surface RNA/glycoRNA in situ, investigating the mechanisms through which glycoRNA is anchored on the cell membrane (I have some evidence that glycoRNA is most likely protected by protein binding) and setting standard for glycoRNAs purification/sequencing. Given that glycoRNAs can be found in many cell types, which is corroborated by our unpublished observations, I speculate that glycoRNAs could play important functions across multiple cell types and in multiple biological settings. Studying cell surface RNA/glycoRNA will open new perspectives for understanding hidden aspects of RNA biology and contribute to RNA society.