1 – Elucidating the Functions of HSF1 Using Domain Specific RNA Aptamers

Lina Bagepalli, Fabiana Duarte, John Lis


The Heat Shock Response is a highly conserved protective mechanism that is regulated at the transcriptional level by transcription factors called Heat Shock Factors (HSFs). These factors, when activated by high temperatures or stress, strongly induce the expression of Heat Shock genes (HS genes), that encode Heat Shock Proteins (HSPs). HSF1 in mammals is the functional homolog of the single HSF in yeast and the fruit fly, and, is required for the rapid activation of HS genes.  HSF1 consists of at least three functional domains; the DNA binding domain (DBD), Trimerization domain (TD) and the Trans-Activation domain (TAD). Each of these domains function to coordinate HSF1’s ability to bind to its target DNA elements and trans-activate HS genes under HS conditions. However, the mechanistic roles of each domain are not entirely understood. To elucidate these functions in living cells, I am using the novel approach of blocking domains with RNA aptamers.

We have thousands of selected aptamers for HSF1, however, their domain specific binding must be characterized in order to test a domain specific effect. I have used two methodologies to characterize the aptamers; Electrophoretic Mobility Shift Assays (EMSA) and a higher resolution, biochemical, UV crosslinking approach.  Presently, I am optimizing the crosslinking approach to characterize the specific regions of HSF1 bound by the RNA. Labeled RNA – HSF1 complexes are UV crosslinked and exposed to RNAse treatment leaving a short, labeled oligonucleotide at every crosslinked site. To identify distinct regions bound by the aptamer, the RNA-Protein complexes are treated with proteases and are run on an SDS PAGE mapping the regions of HSF1 bound by aptamer.

In the future, I will test the effect of using these in vitro characterized domain-specific aptamers as inhibitors by expressing them in vivo and measuring their primary effects on genome-wide transcription using PRO-Seq.

2 – Salmonella Invasion is Controlled through the Secondary Structure of the hilD Transcript

Betteken MI, Hung CC, Eade CR, Nugent SL, and Altier C.


Salmonellosis is estimated to account for one million cases of foodborne illness each year, with Salmonella enterica Typhimurium accounting for the majority. Once within the intestinal tract, the transcriptional regulator of virulence HilD induces Salmonella Pathogenicity Island 1 (SPI1) genes, which begins the invasion process that alters the environment to favor the invading organism’s growth and repress the native microbiota. Regulation of HilD is multi-faceted and involves control at the transcriptional, translational, and post-translational levels. To better understand the regulation of HilD, we undertook a screen to find hilD point mutations that activate SPI1 genes. Interestingly, mutations that resulted in increased expression of SPI1 were found in the 5’ untranslated region of hilD, whose message is predicted to form a stem-loop. We made single-base mutations designed to weaken this secondary structure by mismatching of base pairs, finding them to significantly increase SPI1 expression.  Conversely, replacing an existing A-T pair with a G-C, gaining a single additional hydrogen bond, profoundly reduced expression of these genes. Located within this region is a predicted binding site for CsrA, a translational repressor. Analysis in a truncated CsrA mutant, which does not repress HilD translation, demonstrated increased expression of SPI1 genes in mutants with predicted weakened secondary structure, suggesting that CsrA requires this structure to bind. In contrast, the loss of functional CsrA had no effect in a mutant with greater predicted secondary structure stability, suggesting that CsrA functions as a repressor by stabilizing the structure of the message. These results indicate that Salmonella invasion is finely controlled by the regulation of HilD translation through the interplay of its message with the repressor CsrA.

3 – DNA Damage Responses in Mouse Primordial Germ Cells

Jordana C. Bloom, John C. Schimenti


The ability of organisms to pass their genetic information onwards to subsequent generations is crucial for survival of a species.  In mice, primordial germ cells (PGCs) are set aside early in development to become the germline lineage.  Importantly, while DNA replication associated with rapid cell proliferation is often subject to spontaneous errors, the germline has been shown to be highly refractory to mutation accumulation.  To begin understanding whether PGCs are similar, or dissimilar, to well-studied cell types in their response to altered genomic integrity, I examined cell cycle profiles of PGCs with and without ionizing radiation-induced DNA damage.  At 11.5 days post-fertilization (E11.5), PGCs show no evidence of a G1 cell cycle checkpoint; rather the PGC population and becomes enriched in G2/M phase cells.  While most proliferating cells activate a G1 cell cycle checkpoint in response to DNA damage, mouse embryonic stem cells (ESCs) have also been shown to lack a G1 cell cycle checkpoint in response to DNA damage.  Therefore, we hypothesize that PGCs employ similar cell cycle control and DNA damage response mechanisms to ESCs, and that the latter can be used as an experimental surrogate for investigating genetic quality control in the mammalian germline.

4 – Role of Telomeric Noncoding RNA (TERRA) in Primordial Germ Cell Development and Commitment into Meiosis

Miguel Angel Brieño-Enriquez, Stefannie L. Moak and Paula E. Cohen


Telomeres are dynamic nucleoprotein structures capping the ends of linear eukaryotic chromosomes. They consist of a) Telomeric double strand DNA repeats (TTAGGG); b) Shelterin protein complex; and c) Telomeric Noncoding RNA (TERRA). Telomeres protect the chromosome ends from degradation and erroneous recombination events. The recent discovery of a TERRA as a class of long noncoding RNA (lncRNA) transcribed from the telomeric region has challenged the dogma that telomeres are transcriptionally silenced. Proposed TERRA functions are widespread and include telomere maintenance, telomerase inhibition, telomeric heterochromatin formation, genomic stability, and alternative lengthening mechanism for telomeres. More recently, it has been suggested that TERRA could play an important role in maintenance and/or differentiation of stem cells and Primordial Germ Cells (PGCs). Very little is known about TERRA in the germ line, particularly as it pertains to the viability and integrity of PGCs. We I hypothesize that TERRA expression and/or localization in PGCs is related to gender and gestational age and in consequence the meiotic entry. Using RNA-FISH we localized TERRA in PGCs. RNA-FISH shows the presence of TERRA in both male and female 12.5 and 13.5 days postcoitum (dpc) PGCs as well somatic cells of the fetal gonad. Interestingly, our results show a statistically significant increase in the mean focus count for TERRA with increasing gestational age, indicating that TERRA presence depends on the development stage. In contrast, TERRA focus counts in somatic cells remain constant throughout the time period examined, with high TERRA focus numbers in somatic cells regardless of gestational age. TERRA presence is also sex-specific, TERRA is detectable as early as 12.5dpc in females however in males the first foci appears at 13.5dpc. Expression of TERRA was evaluated using qPCR, showing that the expression of TERRA is both sex- and stage-dependent as well as TERRA expression from different subtelomeric regions.

5 – Dynamic Klf4-mediated Chromatin Reorganization During Somatic Cell Reprogramming

Dafne Campigli Di Giammartino, Andreas Kloetgen, Yiyuan Liu, Daniel Kim, Aris Tsirigos, Effie Apostolu


Cell identity is determined by distinctive gene expression programs, unique epigenetic landscapes, as well as characteristic three-dimensional (3D) genomic organization. Our research focus is to dissect the interplay between TFs and chromatin reorganization during reprogramming from a somatic to an induced pluripotent cell fate. Using Klf4 as a paradigm, we aim to unravel the principles of chromatin reorganization around Klf4-target genes.

We propose that Klf4 does not merely act as a transcription factor to activate or silence genes but also functions as an architectural protein, which is able to drive cell fate change by rearranging chromatin topology from a somatic to pluripotent state.

6 – CRISPR/Cas9 Genome Editing at the Atrial Fibrillation (AF)-associated Pitx2 Locus: The Role of the Long Noncoding RNA Playrr in Cardiac Development and Disease

Frances L. Chen, Ian C. Welsh, John P. Leach, James F. Martin, Natasza A. Kurpios, and Eva M. Oxford


Pitx2 is an essential transcription factor for proper left-right (LR) heart morphogenesis and has maintained postnatal expression within the adult left atrium. Loss of Pitx2 in mice and humans is associated with congenital cardiac defects, increased susceptibility to atrial fibrillation (AF), and development of bilateral sinoatrial nodes (SAN). The genomic mechanisms by which one gene controls multiple crucial developmental milestones with precise spatiotemporal expression are poorly understood. These phenomena likely involve the dysfunctional modulation of Pitx2 by different cis-regulatory noncoding elements. Importantly, the most significantly associated AF genetic loci map to variants in noncoding sequence in an intergenic gene desert near the Pitx2 locus, implicating a role for noncoding elements in the development of AF.

Using transcriptional profiling in mouse and chick, we discovered Playrr, a conserved enhancer-associated long noncoding RNA (lncRNA), located within the Pitx2 locus gene desert. Playrr is expressed asymmetrically (on the right side) and complementary to Pitx2 (on the left) in the developing embryo. To investigate the role of Playrr in regulating Pitx2, we used CRISPR/Cas9 genome editing in mice to target the Playrr RNA transcript while leaving the underlying enhancer intact. RNA in situ hybridization experiments and qRT-PCR revealed that loss of Playrr results in altered Pitx2 isoform expression in embryonic mouse hearts. Additionally, using awake surface ECGs we demonstrate that adult Playrr mutant mice exhibit bradycardia and irregular R-R intervals, indicative of sinus node dysfunction (SND), a bradyarrhythmia and risk factor for the development of AF.  Finally, burst pacing programmed stimulation of Playrr mutant mice revealed that Playrr mutants are predisposed to pacing-induced AF, strikingly similar to Pitx2 heterozygous mice. These data suggest that Playrr modulates Pitx2 gene dosage and isoform expression in the heart leading to conduction abnormalities and predisposing to important cardiac arrhythmias.

7 – Human-specific Changes in Transcription of the Anthrax Toxin Receptor

Lauren A. Choate, Zhong Wang, Charles G. Danko


The bacterium Bacillus anthracis gives rise to anthrax disease by inhibiting the host immune response and causing apoptosis of various cell types. As exposure to anthrax increased in humans, possibly through the advent of agriculture, selective pressures may have affected the human immune system. Signatures of positive selection can be detected between different human populations near the ANTXR2 gene, which encodes a receptor necessary for anthrax toxins to enter cells. We generated Precision Run-On and Sequencing (PRO-seq) and mRNA sequencing data in human, chimpanzee, and rhesus macaque CD4+ T-cells, which reveals an 8-fold decrease in ANTXR2 transcription in humans compared to chimpanzee and rhesus macaque, and a similar downstream effect on mRNA levels. To test if changes in ANTXR2 transcription render human cells more resistant to anthrax than those of other primate species, we are performing viability and activation assays in T-cells after an anthrax toxin challenge in multiple primate species. To find the causal DNA sequence changes that decreased ANTXR2 expression in humans, we are identifying sequence changes at transcription factor (TF) binding sites that are differentially occupied between primate species around ANTXR2. To identify TF binding sites in an unbiased manner, we developed a machine learning tool, dTOX, to classify the binding status of TF motifs based on characteristic patterns in PRO-seq. We used dTOX to identify differences in TF binding activity near the ANTXR2 locus in CD4+ T-cells of human, chimpanzee, and rhesus macaque. Through the intersection of the differential TF binding data with human-specific SNPs near ANTXR2, we have identified several candidate causal sequence changes which may be responsible for the human-specific decrease in transcription.

8 – Chromatin Run-on Reveals Nascent RNAs That Differentiate Normal and Malignant Brain Tissue

Tinyi Chu, Edward J Rice, Gregory T Booth, Hans H Salamanca, Zhong Wang, Leighton J Core, Sharon L Longo, Robert J Corona, Lawrence S Chin, John T Lis, Hojoong Kwak, Charles Danko


Non-coding elements in our genomes that play critical roles in complex disease are frequently marked by highly unstable RNA species. Sequencing nascent RNAs attached to an actively transcribing RNA polymerase complex can identify unstable RNAs, including those templated from gene-distal enhancers (eRNAs). However, nascent RNA sequencing techniques remain challenging to apply in some cell lines and especially to intact tissues, limiting broad applications in fields such as cancer genomics and personalized medicine. Here we report the development of chromatin run-on and sequencing (ChRO-seq), a novel run-on technology that maps the location of RNA polymerase using virtually any frozen tissue sample, including samples with degraded RNA that are intractable to conventional RNA-seq. We used ChRO-seq to develop the first maps of nascent transcription in 23 human glioblastoma (GBM) brain tumors and patient derived xenografts. Remarkably, >90,000 distal enhancers discovered using the signature of eRNA biogenesis within primary GBMs closely resemble those found in the normal human brain, and diverge substantially from GBM cell models. Despite extensive overall similarity, 12% of enhancers in each GBM distinguish normal and malignant brain tissue. These enhancers drive regulatory programs similar to the developing nervous system and are enriched for transcription factor binding sites that specify a stem-like cell fate. These results demonstrate that GBMs largely retain the enhancer landscape associated with their tissue of origin, but selectively adopt regulatory programs that are responsible for driving stem-like cell properties.

9 – miR-181d Regulation of Cytosolic Isocitrate Dehydrogenase 1

Joanna L. Fiddler and Stephen L. Clarke


Characterization of gene expression has allowed for a better understanding of how mammalian systems respond to disease, environmental stressors, and nutrient status.  With the development of microarray and RNA sequencing technologies, microRNA (miRNA) have emerged as a new class of molecular regulators that influence gene expression.  miRNA are small noncoding molecules that posttranscriptionally regulate gene expression by repressing translation or destabilizing mRNA.  Recent evidence suggests miRNA are involved in the molecular coordination of iron homeostasis.  The purpose of this study was to examine miR-181d expression in response to iron deficiency and miR-181d regulation of gene targets.  To characterize the potential role of miR-181d, miRWalk software was used to identify the potential gene target, isocitrate dehydrogenase 1 (Idh1).  In vitro reporter assays demonstrated a site-specific interaction between miR-181d and Idh1 that resulted in a reduction of luciferase activity.  Additionally, overexpression of miR-181d with lentiviral particles resulted in a significant reduction of Idh1 protein levels.  Although, miR-181d was not significantly upregulated in iron deficient rat frontal cortices or iron chelated neuro-2A cells, Idh1 mRNA abundance and protein levels were significantly reduced in both models.  These results demonstrate for the first time that miR-181d posttranscriptionally regulates Idh1 and Idh1 was significantly reduced in response to iron deficiency.

10 – Uncovering Biology of Klinefelter Syndrome (47,XXY) Infertility Using Novel 10x Genomics Single Cell Sequencing

Ryan Flannigan, Ana-Maria Sutii, Fabien Campagne, Jackson Hobgood, Russell Hayden, Peter N Schlegel, Darius Paduch


Finding sperm in rare areas of dilated seminiferous tubules (DSFTs) among the mass of collapsed seminiferous tubules (CSFTs) in men with Klinefelter syndrome (KS) is difficult to explain. Two plausible explanations of sperm production in KS are either spontaneous rescue of 47,XXY and loss of the additional X ch. in some areas of testis, or existence of low grade mosaicism in tissue levels. Our lab has confirmed that all pre-meiotic cells in testis have two X ch. including spermatogonial stem cells (SSCs). Thus, we believe that along SFTs in men with KS there are similar number of SSCs, but due to optimal niches in some regions of testis the rescue of genetic defects is possible. The aim of this study was to characterize heterogeneity of SSCs and somatic cells along the DSFTs and CSFTs from same patient and same testis using single cell sequencing.

We identified markers of SSCs and differentiating spermatogonia (SPG) such as ITGA6, GFRa1, SOHLH2, ENO2, ZBTB16, UCHL1, NEUROG3, OCT4, UTF1, SALL4, CDH1, BCL6B, KIT, ID4 and SOX2/3 in both DSFTs and CSFTs from patients with KS and expressed at similar levels. Thus, indicating that SPGs and SSCs are present in both DSFTs and CSFTs in similar numbers along the SFTs. Presence of SSCs along the SFTs in KS was confirmed with antibody against UTF1 and GFRa1 and normalized to number of Sertoli cells using SOX9. There were statistically significant differences in expression of markers involved in Sertoli to SSCs signaling; thus, confirming that rescue of KS is a result of rare events in which SSCs are exposed to an optimal spermatogenic niche and are able to lose the additional X ch. These findings support our further effort into high resolution 4D analysis along SFTs to pinpoint critical levels of ligands and receptors necessary for KS rescue.

11 – Elucidating Genetic Regulatory Mechanisms Essential for Self-Renewal and Differentiation in Mouse and Human Spermatogonial Stem Cells

Kathryn Grive, Jennifer K. Grenier, Andrew Grimson, Eileen Shu, Paula E. Cohen


Children undergoing chemotherapy for malignancies are often faced not only with the imminent consequences of their disease, but also with the prospects of loss of fertility in adult life. While it is possible to preserve semen samples from adolescent and adult men, pre-pubertal boys are unable to take advantage of this preemptive measure. Though preservation of testis tissue is a potential means of addressing this problem, transplantation of whole tissue involves the risk of re-seeding the cancer cells. Therefore, isolation and in vitro differentiation of spermatogonial stem cells (SSCs), the self-renewing stem cells which serve as the progenitors for the entire sperm population, presents a unique opportunity to help preserve the fertility of young men whose natural complement of germ cells is lost due to chemotherapeutics.

While SSC isolation can now be performed robustly using sedimentation and magnetic cell sorting protocols, global gene regulation remains to be comprehensively studied in these cells. To this end, high-throughput sequencing was utilized to identify genes which are highly expressed in isolated SSCs relative to isolated spermatogonia. From this analysis, we have identified enrichment of transcripts from known SSC markers including Plzf, Gfra1, and Sall4, as well as potentially novel markers of the SSC population. Furthermore, small RNA-sequencing analysis of these samples was performed to better understand the unique roles that miRNAs and piRNAs may play in the regulation of this cell population essential for fertility. We have now identified several highly SSC-enriched miRNAs, as well as miRNAs differentially expressed in isolated spermatogonia. We have further expanded our analysis to include isolated Integrin a6-positive human SSCs, and have identified canonical genetic signatures in these cell types. In collaboration with our RNA-Sequencing Core, we are currently comparing SSC-enriched miRNAs between mice and humans to identify potentially conserved genetic regulatory mechanisms in the mammalian SSC population, as well as validating expression and targeting of these potential regulatory RNAs.

These studies are providing greatly enhanced knowledge of the regulation of spermatogenesis, but will also a potential basis for clinical applications utilizing these methods for fertility preservation. Equally as important is refining our understanding of the mechanisms underlying the production of healthy sperm, which can then participate in fertilization to create healthy embryos.

12 – Microfluidic Full-length mRNA Sequencing Reveals Complex Splicing Patterns in Alzheimer’s Disease Candidate Genes

Ishaan Gupta, Fereshteh Jahanbani, Mortein Rasmussen, Michael Snyder, Hagen Tilgner


RNA sequencing has revealed splicing to be prevalent in the majority of mammalian genes. It is estimated that splicing generates on average 10 or more different transcriptional products (or isoforms) per gene, with each transcript having a specific combination of exons. However, due to short read lengths in RNA-sequencing, the current view of gene expression is limited to studying each splicing event in insolation and we are unable to obtain an accurate quantitative abundance of the complete transcript isoforms. Moreover recent studies, show that two isoforms of the same gene can encode varied protein-structures that interact with different set of proteins and perform very different functions. Therefore, in order to understand the exact biological output of a gene it is imperative to look at the entire molecular output of transcription (“what genes are saying”), not just its abundance (“how much genes are talking”).

We have developed a droplet-based microfluidic technique that enables quantification of more than 10 million distinct full-length mRNA molecules to reveal the exact juxtaposition of exons in a transcript. Our method obtains these molecules experimentally unlike the prevalent reconstruction-based methods that rely on short read paired-end sequencing and quantify 30 times more RNA molecules than prevalent long-read sequencing technologies (PacBIO or Oxford Nanopore). Assaying mRNA from human brain reference, we found specific non-random patterns of arrangement of distant exons for transcripts of the BIN1 and MAPT, top Alzheimer’s disease candidate genes. We found non-random splicing to be much more prevalent (and specific to coding genes) than previously appreciated and estimate that it may act on a majority of the genes to diversify their functional potential. Our results motivate the adoption of long-read RNA sequencing methods to fully understand the functional potential of genomes.

13 – Discovery of The Molecular Mechanisms That Underlie Microrna-29 Mediated Control of Lipogenesis in The Liver

Y-H Hung, C Kurtz, M Kanke, P White, M Deshmukh, M Hussain, X Li, P Sethupathy


MicroRNAs (miRNAs) are important regulators of diverse metabolic pathways. Evidence suggests their roles as etiological factors and/or potential therapeutic targets for many diseases including type 2 diabetes (T2D) and hyperlipidemia. Previously, we reported elevation of miR-29 in the liver of rodent models of obesity or overt T2D. Furthermore, we demonstrated that inhibition of miR-29 significantly lowers circulating lipids in chow-diet fed wild type (WT) C57BL/6J mice. However, the molecular mechanisms underlying this lipid-lowering effect remains unclear. We hypothesized that miR-29 controls lipid synthesis in the liver in part through regulation of Sirtuin-1 (Sirt1), which we have previously reported as a putative target of miR-29, and which others have shown is involved in the suppression of hepatic lipid synthesis. In the present study, we first showed that miR-29 is aberrantly elevated in the liver of Zucker Fatty obese rats compared to lean littermates. Next, we demonstrated that suppression of miR-29 by systemic administration of locked nucleic acid (LNA) inhibitors (20 mg/kg) in chow-fed WT C57BL6/J mice leads to a robust reduction in de novo hepatic lipid synthesis. To determine whether this result is dependent on Sirt1, we repeated the experiment in mice with Sirt1 liver-specific knock-out. Surprisingly, we found that hepatic Sirt1 is not essential for the lipid-lowering effect. However, the results show that hepatic levels of Energy Homeostasis-Associated Protein (Enho), a gene that harbors a predicted miR-29 target site, are significantly upregulated by LNA treatment and that this response is intact in the absence of Sirt1. We confirmed that miR-29 regulates Enho through transgenic over-expression of miR-29 in mice, and also established that miR-29 directly targets Enho through 3’ UTR reporter gene assays. Overall, the findings uncover the potential centrality of hepatic Enho in the molecular network through which miR-29 regulates hepatic lipid synthesis and possibly systemic insulin sensitivity.

14 – m6A RNA Methylation Drives Liquid-Liquid Phase Separation and Mediates Stress Granule Formation

Pierre Klein, Ryan Ries, Sara Zaccara, Brian Pickering, Deepak Patil, Samie Jaffrey


m6A is the most abundant internal modification in mRNAs.  It recruits m6A-binding proteins, of which the most prominent are the YTH-domain containing proteins.  Among these proteins, the YTHDF family – YTHDF1, 2 and 3 – are the main cytoplasmic readers.  These proteins are highly similar and contain the so-called YTH domain, which binds m6A, together with a low complexity domain, an unstructured region that contains low amino-acid diversity.  Although YTHDF proteins have been intensively studied, their function is still controversial and the role of the low complexity domain is poorly understood. Low complexity domains are found in many RNA-binding proteins and promote liquid-liquid phase separation.  This physical phenomenon explains how two miscible liquids demix, resulting in two separate phases. Liquid-liquid phase separations are fundamental in cell biology, since they enable the formation of membrane-less domains such as RNA granules, P-bodies, and stress granules.  Stress granules are composed of RNA and RNA-binding proteins with low-complexity domains.  Recent proteomic studies have revealed that YTHDF proteins are a component of the stress granules.  These findings raised the possibility that stress granule formation, which is a critical response to cellular stress, is fundamentally linked to m6A mRNA. We studied the role of YTHDF proteins in stress granule formation and we discovered that 1) YTHDF proteins are able to undergo liquid-liquid phase separation and this requires binding to m6A clusters; 2) YTHDF proteins relocalize to stress granules after cellular stress in an m6A dependent manner; 3) Stress granule formation is markedly impaired in m6A-deficient cells; and 4) YTHDF proteins are required for stress granules formation. Together these results fundamentally advance our understanding of stress granule formation and broaden our understanding of m6A and its role in cellular stress.

15 – Genetic Analysis of Enhancer RNA (eRNA) Variation in Human Population

Katla Kristjánsdóttir, Hyun Min Kang, Hojoong Kwak


Enhancer RNAs (eRNA) are non-coding RNAs transcribed bidirectionally from enhancers. The amount of eRNA generated is directly related to enhancer activity and proposed to be a better predictor of regulatory regions than chromatin modifications or accessibility. Therefore, population-scale analyses of eRNAs will provide a more comprehensive view of genetic variations that regulate gene expression. Quantifying eRNA levels has been challenging due to their instability. However, nascent RNA sequences can capture and quantify eRNAs under active transcription. Using PRO-cap in lymphoblastoid cell lines from 67 Yoruba individuals, we quantified eRNA levels to identify actively transcribed enhancers in high resolution. We found 74,988 enhancer candidates in LCLs. These enhancers have a specific structure, composed of central transcription factor binding sites flanked by divergent bidirectional eRNA initiation sites. The enhancers are enriched with known expression quantitative trait loci (eQTLs) and disease associated variants. Furthermore, we identified thousands of enhancers where eRNA levels are significantly (FDR < 0.05) associated with genotypes (transcription initiation quantitative trait loci: tiQTLs) or where genotypes affect the relative ratio between the bidirectional eRNAs (directional initiation quantitative trait loci: diQTLs). Interestingly, while tiQTLs are enriched in the central transcription factor binding sites, diQTLs are enriched near the eRNA initiation sites, reflecting a dual-hub model of enhancer architecture. At clustered enhancers, this architecture affects interactions between the enhancers. Correlation analysis between enhancers, and between enhancers and promoters, shows that interactions are highly enriched within 200 kilobases, and disrupted by intervening insulators or strong promoters. Our results collectively provide evidence that genetic variation of enhancer function relies not only on transcription factor binding, but also on eRNA transcription itself.

16 – Post-transcriptional Regulation of De Novo lipogenesis by mTORC1-S6K1-SRPK2 Signaling

Gina Lee, Yuxiang Zheng, Sungyun Cho, Cholsoon Jang, Long He, Paola M. Cavaliere, Andre Chavez, Noah E. Dephoure, Joshua D. Rabinowitz, Lewis C. Cantley, John Blenis


mTORC1 is a signal integrator and master regulator of cellular anabolic processes linked to cell growth and survival. Here we demonstrate that mTORC1 promotes lipid biogenesis via SRPK2, a key regulator of RNA-binding SR proteins. mTORC1-activated S6K1 phosphorylates SRPK2 at Ser494, which primes Ser497 phosphorylation by CK1. These phosphorylation events promote SRPK2 nuclear translocation and phosphorylation of SR proteins. Genome-wide transcriptome analysis reveals that lipid biosynthetic enzymes are among the downstream targets of mTORC1-SRPK2 signaling. Mechanistically, SRPK2 promotes SR protein binding to U1-70k to induce splicing of lipogenic pre-mRNAs. Inhibition of this signaling pathway leads to intron retention of lipogenic genes which triggers nonsense-mediated mRNA decay. Genetic or pharmacological inhibition of SRPK2 blunts de novo lipid synthesis, thereby suppressing cell growth. These results thus reveal a novel role of mTORC1-SRPK2 signaling in post-transcriptional regulation of lipid metabolism and demonstrate that SRPK2 is a potential therapeutic target for mTORC1-driven metabolic disorders.

17 – An RNAi Screen to Identify Novel Genes Involved in the miRNA Pathway

Jacob Merle, Andrew Grimson


MicroRNAs (miRNA) are small (~22-nt) endogenous noncoding RNAs, which regulate gene expression through binding complementary target sites in mRNAs.  To exert their regulatory function, primary miRNA transcripts must undergo processing steps in both the nucleus and cytoplasm, before the mature miRNA can assemble into the miRNA-induced silencing complex (miRISC).  Once the miRISC assembles, the miRNA acts as a guide, directing translational repression and decay of mRNA targets. Although the core components of the miRNA biogenesis and effector pathways are known, there are multiple gaps in our knowledge of miRNA biogenesis and function. For example, many questions remain regarding the mechanism by which miRNAs are able to direct translational repression and accelerated mRNA decay, and most current models suggest the existence of additional, currently unknown, components of the silencing pathway.  Identifying additional genes involved in miRNA mediated repression will help us answer some of these outstanding questions.
We are developing a high throughput method to screen a genomic short hairpin RNA (shRNA) library for knock-down events which alter miRNA-mediated repression. To identify novel genes involved in the miRNA pathway, we have developed a reporter cell line that is sensitive to RNAi events that alter miRNA-mediated repression, and have recently performed a fluorescence-based RNAi screen using a small shRNA library. Utilizing fluorescence activated cell sorting (FACS) and high throughput sequencing we compared cells with altered reporter activity to an unsorted background. We observed the enrichment of shRNAs targeting known miRNA pathway components which validates this approach and also identified candidate novel regulators/effectors of miRNA-mediated repression.

18 – The Relationship Between Chromatin Conformation and Transcription Under the Induction of Heat Shock

Paul R. Munn (Cornell University); Judhajeet Ray (Cornell University); Abdullah Ozer (Cornell University); Anniina Vihervaara (Cornell University); Fabiana M. Duarte (Cornell University); John T. Lis (Cornell University); Charles G. Danko (Cornell University)


Chromatin folding within the interphase nucleus is involved in the control of gene regulation. However, how chromatin conformation is affected following the induction of external stressors, such as heat shock, is not well understood. To understand the link between changes in chromatin architecture and transcription, we conducted the first integrative comparison between primary transcription (by PRO-seq), transcription factor binding (by ChIP-seq), and chromatin conformation (by Hi-C) following short durations (20 min.) of heat shock in Drosophila melanogaster S2 cells. Using both computational and visual examination we could find no evidence for genome-wide changes in Hi-C contact frequency following heat shock, even at loci with extensive transcriptional differences. This contradicts previous reports that short durations of heat shock affect chromatin structure in Drosophila Kc cells. We specifically examined chromatin interactions at a subset of genes whose up-regulation is dependent on heat shock factor (HSF), but with no detectable promoter proximal HSF binding by ChIP-seq. These HSF-dependent genes exhibit closer contact frequencies with nearby HSF binding sites than expected by a genome-wide null model both pre- and post- heat shock.  However, we found no evidence for heat-shock dependent changes in contact frequencies between these HSF-dependent gene promoters and nearby HSF binding sites. Our results suggest that the chromatin loops necessary to bring HSF binding sites and the promoters of heat shock genes into close proximity, and thus facilitate a response to the induction of heat shock, are established in advance of the stimulus, and remain unchanged after the stimulus has been applied.

19 – How Type II CRISPR-Cas Systems Establish Prokaryotic Adaptive Immunity Through Cas1-Cas2 Mediated Spacer Integration

Yibei Xiao, Sherwin Ng, Ki Hyun Nam, Ailong Ke


Clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated (Cas) systems provide adaptive immunity in prokaryotes by acquiring fragments of invading nucleic acids and integrating them as spacers into the CRISPR array. In the type II-A CRISPR system, spacer integration is facilitated by a complex of the highly conserved Cas1 and Cas2 proteins. Using X-ray crystallography, we obtained three structural snapshots of the Enterococcus faecalis type II-A Cas1-Cas2 integrase complex in four functional states during spacer integration.

Our biochemical analysis determined that EfaCas1-Cas2 preferentially binds to a 30-bp dual-forked prespacer with 4-nt 3’-overhangs. After prespacer binding, Cas1 stochastically searches for the half site containing the leader and inverted repeat sequence on the CRISPR array through tilting of the target DNA over a Cas2 fulcrum. Target recognition in the minor groove of the leader sequence triggers the prespacer terminal 3’OH to undergo a nucleophilic attack at the leader-repeat junction, ligating one strand of the CRISPR array to the 3’-overhang of the prespacer. Notably, target recognition is facilitated by DNA bending, a prerequisite for full integration into the CRISPR array.

The snapshots of Enterococcus faecalis Cas1-Cas2 1) loading a prespacer, 2) sampling half-sites, 3) promoting leader-proximal half-site integration, and 4) completing full-site integration elucidate the entire mechanism for spacer integration in the type II CRISPR-Cas system, allowing for potential applications in genome barcoding, information storage, and lineage tracing.

20 – Disentangling MicroRNA Regulation: Differentiating Direct and Indirect Targets

Ravi K Patel, Jessica West, Andrew W Grimson


MicroRNAs (miRNAs) repress the expression of many mRNAs (direct targets) via complementary binding of the miRNA seed to target sites within 3′ untranslated regions (UTRs). The post-transcriptional repression of direct targets consequently alters the expression of other downstream genes (indirect targets), which are likely to be transcriptionally regulated. Together, these direct and indirect targets form a complex regulatory network that defines miRNA’s roles in cells. Hence, studying the miRNA-regulatory networks by differentiating direct and indirect targets is essential to characterize functions of miRNAs. Despite extensive research in this direction, the current methods to identify targets suffer from high false positive rates and are inadequate to distinguish direct from indirect targets. We developed an experimental strategy to identify and distinguish direct targets (mRNAs regulated at post-transcription) from indirect targets (secondary effects at transcription) by subtracting transcriptional changes (measured using PRO-Seq) from changes in steady-state mRNA levels (profiled using RNA-Seq). We tested the efficacy of this approach using stable cell lines ectopically expressing miRNAs, miR-1 and miR-122, upon induction. We found that our approach identifies direct and indirect targets of used miRNAs with high specificity, which we also confirmed by Argonaute-CLIP-Seq experiments. While the majority of identified direct targets exhibit matches to miRNA-seed sequence in their 3′ UTR, our approach also identifies atypical targets containing other classes of sites. The network reconstruction by combining direct and indirect targets indicated that miR-1 acts a master regulator, whereby miR-1 represses a handful of transcription factors and RNA binding protein to elicit wide-ranging effect on the transcriptome via extensive indirect targeting. Interestingly, miR-122 regulates relatively small number of mRNAs primarily via direct targeting. We are currently extending this approach to study in vivo functions of endogenous miRNAs in mice. This new approach provides a powerful method to reveal miRNA regulatory networks with precision.

21 – Redox-Responsive Post-Transcriptional Regulation of the Nrf2/Antioxidant Response Pathway

Jesse R. Poganik, Alexandra Van Hall-Beauvais, Marcus J.C. Long, Michael Disare, Yi Zhao, Xuyu Liu, and Yimon Aye


The Nrf2-antioxidant response pathway is indispensable for the cellular response to oxidative and electrophilic stress.  The growing importance of this pathway is underscored by the recent FDA approval of the drug Tecfidera, which is believed to function at least in part through activation of Nrf2 signaling. However, the majority of our understanding of how activation of this critical pathway is controlled is currently limited to protein-level regulation of Nrf2. We recently discovered novel mRNA-level regulation of Nrf2 by two mRNA-binding proteins, both of which possess important connections to various disease states. Interestingly, one of these proteins was discovered to be a privileged sensor of the bioactive endogenous signaling electrophile 4-hydroxynonenal (HNE). We are currently exploiting our lab’s recently developed T-REX electrophile delivery platform to understand how the HNE-modified state of this protein manifests distinct regulatory activity relative to the unmodified state. Our data highlight a novel, dynamic, redox-linked regulatory mode of Nrf2 which we hope will ultimately shed more light on the regulation of this essential stress-response pathway.

22 – A Comparative Stem Cell Model to Study the Molecular Basis for the Natural Variation in Mammary Cancer Development

Gat Rauner, Melissa M. Ledet, and Gerlinde R. Van de Walle


It has long been observed that some species rarely develop malignancies of the mammary gland and are virtually resistant to mammary cancer. The mechanisms behind this phenomenon are largely unknown. Mammary cancer is thought to originate from mammary stem and progenitor cells (MaSC), which are tissue-resident, long-lived cells with regenerative capacity. MaSC have been extensively studied in humans and in mice, less so in other mammals.

We have been isolating and culturing MaSC collected from fresh mammary tissue of a variety of mammalian species, and are employing a species-comparative approach in order to identify mechanisms that enable MaSC in mammary cancer-resistant species to evade malignant transformation.

As part of this approach, we compared global gene expression in MaSC from dogs (susceptible) and horses (resistant) using RNAseq. While the results are still being analyzed, interesting findings have already emerged. For example, the expression of positive regulators of the oncogene MDM2 (AKT1, NFATC1, SP1) were higher in canine MaSC, whereas negative regulators of MDM2 (ATM, CDKN2A, GORAB) were higher in equine MaSC. MDM2 is a potent negative regulator of p53, indicating the potential role of the p53 pathway in mediating mammary cancer resistance in horses.

To obtain a more comprehensive view of differences in gene expression and regulation, we are planning to compare the expression of microRNA (miRNA) in MaSC from resistant and susceptible species. miRNAs function as post-transcriptional regulators of gene expression, with established roles in breast cancer. They are particularly interesting for cross-species comparison, as they represent a combination of deep sequence conservation and rapid evolution. These experiments will be done in close collaboration with Dr. Sethupathy.

Studying genes and miRNAs that are differentially expressed between mammary cancer-resistant and -susceptible species will increase our knowledge of mammary cancer biology and will help define regulatory mechanisms that impact breast cancer risk.

23 – miR-375 Controls Intestinal Stem Cell Proliferation

Ajeet P. Singh, Michael T. Shanahan, Matt Kanke, Vera D. Rinaldi1, Yu-Han Hung, Bailey C. E. Peck, Ennessa G. Curry, John C. Schimenti, and Praveen Sethupathy


MicroRNAs (miRNAs) have emerged as molecular sensors of environmental stimuli and adaptive responders to diverse perturbations. The intestine faces a barrage of constantly changing stimuli, including luminal microbes and dietary factors. To accommodate the high functional demand, the intestinal epithelium renews every 4-5 days – process driven by intestinal stem cells (ISCs), which are responsible for generating all of the differentiated cell types of the intestinal epithelium. Recent studies have shown that high-fat diet (HFD) increases the proliferation of ISCs, and alters intestinal morphology and physiology, but the molecular mechanisms remain poorly characterized. We hypothesized that miRNAs in ISCs are sensitive to HFD and that they contribute to shaping the intestinal response to HFD. In this study, we set out to test this hypothesis. We first demonstrated through small RNA sequencing that several miRNAs are robustly expressed in ISCs according to four independent FACS-based methods of ISC isolation, including sorting of Lgr5+ cells. We found that miR-375 in particular is highly enriched in ISCs relative to the other cell types of the IE, except for enteroendocrine cells where its expression is the highest. We then showed that a 5-month obesogenic HFD, which boosts ISC proliferation, leads to a significant reduction in the levels of miR-375 in ISCs. Moreover, integrative transcriptomic analysis revealed that miR-375 is a candidate master regulator of genes that are significantly up-regulated by HFD. Robust suppression of miR-375 in 3-D mouse enteroids established from the jejunum of wild-type, Sox9-EGFP, or Lgr5-EGFP reporter mice, led to a hyper-proliferative phenotype. Data from genetic deletion studies showed that miR-375-/- mice exhibit significant increases in jejunal crypt proliferation. The next steps are to define the key downstream targets of miR-375. Our study has identified critical miRNAs in the response of the intestinal crypt to an obesogenic diet.

24 – Alternative ATF4 Translation Under Sulfur Amino Acid Deprivation

Robert Swanda, Jun Zhou, & Shu-Bing Qian


Amino acids are building blocks for protein synthesis. In response to the shortage of amino acids, cells rapidly activate a pathway called amino acid response (AAR) that is essential for cell survival. In eukaryotic cells, the AAR involves the activation of general control nonderepressible 2 (GCN2) kinase that phosphorylates the alpha subunit of eukaryotic initiation factor 2 (eIF2α) at serine residue 51. By limiting the delivery of initiator tRNA, phosphorylated eIF2α suppresses global protein synthesis. However, AAR also triggers the translation of activating transcription factor 4 (ATF4) via an alternative initiation mechanism. ATF4 regulates the expression of several amino acid biosynthetic genes as well as their transporters to ensure intracellular metabolic homeostasis. The traditional view of AAR is based on full amino acid starvation, but very little is known about the cellular response to individual amino acid deprivation. We wanted to determine whether the similar ATF4 response could be induced by individual amino acid deprivation. We took advantage of mouse embryonic fibroblasts (MEFs) bearing wild type (S/S) or non-phosphorylatable eIF2α mutant (A/A) exposed to time course single amino acid deprivation. Our results show varied ATF4 expression levels during individual amino acid deprivation in S/S cells at both the transcriptional & translational levels, and unexpectedly A/A cells also experienced the increased ATF4 levels under particular amino acid deprivations. These findings are supported when examining GCN2 -/- cells and through the use of a small molecule modulator, ISRIB. We have begun to identify a complex signaling pathway involving oxidative stress, antioxidants, and amino acid transporters. This pathway, under control of Nrf2, may be driving this alternative mechanism primarily through changes in xCT and GHS. Given the broad functionality of amino acids, there is an urgent need to understand how individual amino acid deprivation alters gene expression and protein homeostasis.

25 – Biological and Chemical Probing of RNA Structures in a Positive Sense RNA Virus

Kyle E. Watters, Krishna Choudhary, Sharon Aviran, Julius B. Lucks, Keith L. Perry and Jeremy R. Thompson


In single stranded (+)-sense RNA viruses, RNA structural elements (SEs) play essential roles in the infection process from replication to encapsidation. Using selective 2’-hydroxyl acylation analyzed by primer extension sequencing (SHAPE-Seq) and covariation analysis, we explore the structural features of the third genome segment of cucumber mosaic virus (CMV), RNA3 (2216 nt), both in vitro and in plant cell lysates. Comparing SHAPE-Seq and covariation analysis results revealed multiple SEs in the coat protein open reading frame and 3’ untranslated region. Four of these SEs were mutated and serially passaged in Nicotiana tabacum plants to identify biologically selected changes to the original mutated sequences. After passaging, loop mutants showed partial reversion to their wild-type sequence and SEs that were structurally disrupted by mutations were restored to wild-type-like structures via synonymous mutations in planta. In addition to SE structural reversions, there were for some disrupted SEs associated in planta mutations that caused distal amino acid changes in specific structural regions of the coat protein. These results support the existence and selection of virus open reading frame SEs in the host organism and provide a framework for further studies on the role of RNA structure in viral infection. Additionally, this work demonstrates the applicability of high-throughput chemical probing in plant cell lysates and presents a new method for calculating SHAPE reactivities from overlapping reverse transcriptase priming sites.

26 – Global Small RNA-mediated Regulatory Networks in Arabidopsis thaliana

Jose Vargas, Keith Perry


sRNAs are important regulators of gene expression in many eukaryotic systems, and several studies have proposed they function as ‘master regulators’.  In plants, there is increasing evidence for a networked mode of action for sRNAs. Regulation involves information cascades derived from a single microRNA-transcript targeting event, leading to the production of phased interfering RNAs (phasiRNAs) with the potential to regulate gene expression in cis and trans.  Information cascades are expanded by the phasiRNA-triggered production of secondary and tertiary phasiRNAs targeting additional transcripts.  Genomic level studies of sRNA-mediated networks have thus far been limited to in silico analyses wherein the biological relevance has not been assessed.  The objective of this study was to identify and characterize biologically relevant sRNA-mediated regulatory networks at a genome-wide level.  This was achieved by producing corresponding sRNA-seq, RNA-seq and degradome datasets for wild type A. thaliana.  A computational pipeline was developed using current databases and phasiRNA biogenesis features to construct a meta-network. Using this approach, we doubled the number of biologically active phasiRNAs identified compared to existing tools. Due to our relatively large degradome dataset, the resulting meta-network includes the most comprehensive analysis of sRNA (miRNA and phasiRNA) cleavage activity to date. Enhanced identification of sRNAs and their targets allowed us to produce the largest sRNA-mediated regulatory network of biological relevance to date. Composition analysis indicated that approximately 70% of all A. thaliana known genes are under sRNA regulation; functional analysis showed that around 29% of all annotated regulatory genes are represented in the network. We are using this meta-network to better understand structural and dynamic changes in the host sRNA-mediated regulatory response to pathogen infection.

27 – A Novel Methodology to Modulate Immune Environment Using Synthetic mRNA

Yitian Xu1, Lu Huang2, Jonathan L. Kirschman3, Daryll Vanover3, Pooja M. Tiwari3, Philip J. Santangelo3*, Xiling Shen1,4,5*, David Russell2*
1. Department of Biological and Environmental Engineering, Cornell University, Ithaca, NY, USA.
2. Department of Microbiology and Immunology, Cornell University, Ithaca, NY, USA.
3. Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, Georgia.
4. School of Electrical and Computer Engineering, Cornell University, Ithaca, NY, USA.
5. Department of Biomedical Engineering, Duke University, Durham, NC, USA.
*: co-corresponding authors


As the next generation of gene therapy, synthetic mRNA therapy is known for its intrinsic advantages to avoid genomic integration that associated with DNA-based therapy, and its prolonged production of target proteins over traditional recombinant protein therapy. Synthetic mRNA therapy has been studied in various disease models including cancer, infectious disease and cardiovascular disease. However, these studies focus mainly on expressing proteins with direct therapeutic effect or antigens to elicit an immune response. It is not well established that whether synthetic mRNA can be taken advantage in cell based therapies in vivo to recruit cells of interest and modify them locally. Here we reported a synthetic mRNA based immune-modulation method, exampled by using ccl2 and ccl3 synthetic mRNA to recruit a certain population of monocyte in a non-inflamed manner. These recruited monocytes exhibit neither bactericidal or tissue-repairing behaviors, but stay in a neutral, non-programmed state. With additional IFN-gamma mRNA and IL-4 mRNA, these cells can be polarized to different phenotypes. Furthermore, the monocytes recruited by ccl3 and IFN-gamma mRNA are able to launch the most rapid and strong superoxide burst, which demonstrates their active functionality upon proper signals. In summary, our data established a synthetic mRNA based immune-modulation system, which allows recruitment and modification of specific immune cell populations in a precise and localized manner, and more importantly, proves the concept that synthetic mRNA can be utilized in cell based therapies to replace the time-consuming and labor-demanding process of modifying patients’ own cells ex vivo and transferring back to patients afterwards.

28 – Time-resolved SAXS and Ensemble Modelling Reveal Mg Orchestration Across an RNA Folding Landscape

Alex Plumridge, Andrea Katz, George Calvey, Ron Elber, Serdal Kirmizialtin and Lois Pollack


RNA’s unique folds and motions enable many of its newly discovered functions, yet RNA self-assembly and conformational dynamics are not well understood. Here, we use microfluidic mixing to trigger the Mg induced folding of tP5abc; a representative of the common RNA motif the three helix junction. Time-dependent conformations are measured using time-resolved small angle x-ray scattering (SAXS), and analyzed by pairing ensemble decomposition with all-atom simulations. Detailed structural ensembles, derived at crucial stages of folding, reveal distinct phases of tP5abcs journey to the native state. Many of these phases rely on the presence of Mg ions, and underscore their unique and varying role(s) in guiding RNA through a complex and rugged landscape.

29 – The RNA Sequencing Core at Cornell University

Christine A. Butler, Jennifer K. Grenier


The RNA Sequencing Core (RSC) provides high quality gene expression profiling services to the research community at Cornell. The RSC offers fully integrated package services that include library preparation, sequencing, and standard analysis for two modes of RNA analysis: transcriptome profiling (RNAseq) and small-RNA profiling.

For RNAseq, the RSC generates libraries using either mRNA (polyA+ RNA) or ribosomal RNA depleted RNA as input. The rRNA-depletion option provides the opportunity to profile non-polyadenylated RNA transcripts or to utilize degraded RNA samples that historically have been difficult to profile with standard methods. RNAseq libraries are sequenced to a minimum target depth of 20M reads per library, and standard analysis includes raw data QC, preprocessing, transcriptome/genome mapping, and differential gene expression analysis using the Tuxedo package.

For small RNA sequencing, the RSC generates libraries of small RNAs (18-32nt) using a protocol that exploits their distinct terminal chemistry and size. We have generated small RNA datasets from cellular total RNA as well as cell-free RNA extracted from plasma and extracellular vesicles. Small RNA libraries are sequenced to a minimum target depth of 10M reads per library and analyzed with miRDeep2 software to generate normalized counts for annotated mature microRNAs.

The RSC also can provide specialized RNAseq services such as analysis of multiple organisms from a single sample (dual RNAseq), pathogen discovery, and transcriptome discovery/assembly, following consultation with and approval by the Core Director. In the future, RSC services may include RNAseq for ultra-low input samples (10-1000 cells) that require RNA amplification and ATACseq to identify regions of accessible chromatin.

We work closely with researchers to advise on experimental design and RNA input/quality recommendations in order to maximize data quality and utility. Our goal is to partner with discovery researchers and deliver high quality gene expression profiling results to enable their research goals.

30 – Biotechnology Resource Center

VanEe, J., Pillardy, J., Schweitzer, P., Williams, R., Zhang, S., Rose, J.


The mission of the Cornell University Biotechnology Resource Center (BRC) is to catalyze and promote research in the life sciences.  The BRC provides advanced technologies, services, training and education to the university community and to outside investigators, operating in both fee-for-service and collaborative modes.  The BRC’s core facilities and expertise include genomics, proteomics, metabolomics, imaging, bioinformatics.