Skirball Institute of Biomolecular Medicine Doctoral Thesis Defenses
After completing their scientific training, doctoral candidates working at NYU Langone’s Skirball Institute of Biomolecular Medicine discuss the significance of their original research at a thesis defense. Members of the public are welcome to attend.
Doctoral Thesis Defenses 2019–20
Congratulations to everyone who defended!
Lauren Bayer Horowitz—Ringstad Lab
Title: “Molecular Mechanisms of p38 MAP Kinase Signaling During Neuronal Differentiation”
Date: November 1, 2019
Abstract: The nervous system comprises diverse and highly specialized neuron types, each expressing a unique set of genes that defines its functional properties. What molecular mechanisms generate this diversity remains a central question in neuroscience. The nematode C. elegans is a powerful model to study this question. Although small, the C. elegans nervous system displays a high degree of diversity and uses conserved factors to differentiate. Study of a pair of chemosensory neurons, the BAG neurons, showed that a p38 MAP kinase (MAPK), PMK-3, is required for their proper differentiation. p38 MAPKs function in many neuronal processes, but their role in nervous system development is poorly understood.
Here I find that PMK-3 promotes the differentiation of BAG cells by limiting an insulin-like peptide (ILP) signal that represses expression of a BAG neuron fate. pmk-3 mutant BAG cells over-express ILPs, which are released from developing BAGs in an activity-dependent manner. These ILPs act as autocrine signals and initiate a non-canonical insulin-signaling pathway. My findings reveal an unexpected role for insulin signaling in nervous system development, and suggest that insulin-like factors integrate intrinsic genetic programs and neural activity to regulate neuronal differentiation. Because the molecular constituents of this pathway are all highly conserved, I propose that p38 MAPKs and insulin signaling might play similar roles during development of vertebrate nervous systems.
I have also characterized 13 additional mutations that modify the gene expression defects of pmk-3 mutant BAG cells and identified a new complementation group. The affected gene encodes a large, secreted protein with conserved structural motifs but whose molecular function remains unknown. This gene might be a novel component of the autocrine insulin signaling pathway regulated by p38 MAPKs or it might function in a parallel mechanism that regulates neuronal differentiation.
Cynthia Chen—Schwab Lab
Title: “Deciphering the Role of Integrins in T Cell Acute Lymphoblastic Leukemia Infiltration of the Central Nervous System”
Date: November 1, 2019
Abstract: T cell acute lymphoblastic leukemia (T-ALL) has a predilection for the central nervous system (CNS). However, it is unclear how T-ALL enters the CNS. The molecular mechanisms that regulate normal T cell migration across the endothelial blood-brain barrier have been extensively studied and effectively targeted. Upon encountering inflamed endothelium, activated T cells participate in a canonical capture/rolling, arrest, and diapedesis cascade. Two integrins, VLA-4 (α4β1) and LFA-1 (αLβ2), are critical in this cascade for T cells to cross the barrier.
Here, we show that VLA-4 and LFA-1 are dispensable for entry of T-ALL into the CNS. Strikingly, loss of these integrins provided T-ALL cells an advantage in accumulating in the CNS compared to control T-ALL. I confirmed loss of VLA-4 and LFA-1 was directly responsible for T-ALL accumulation in the CNS both by restoring these integrins in a double-knockout line and by knocking out these integrins in an independent T-ALL line. I am exploring possible mechanisms behind the accumulation of VLA-4 and LFA-1 deficient T-ALL in the CNS, including increased homing and survival, as well as decreased exit from the CNS via dural lymphatics.
Our results provide an unprecedented clue to the trafficking of T-ALL into the CNS and show that the migratory requirements of normal T cells and leukemic cells are distinct, a feature that may be exploited for their differential targeting in treatment. Our findings also raise caution about proposals to use integrin blockade in conjunction with chemotherapy for leukemia to dislodge cells from their niche in the bone marrow or to overcome chemotherapy resistance.
José G. Montoyo-Rosario—Nance Lab
Title: “The Role of PKC-3/aPKC and Genetic Suppressors in C. elegans Epithelial Cell Junction Formation”
Date: August 28, 2019
Abstract: Epithelial cells, which line our organs, possess adhesive junctions that enable them to function as a barrier to the outside environment. Loss of junctions compromises tissue structure and has been linked to human diseases such as kidney disease, cancer, and birth defects. Junctions form when clusters of junction proteins concentrate at cell contacts near the apical surface then coalesce into belt-like structures. The atypical protein kinase C (aPKC), which functions together with the scaffolding protein PAR-6, is required for junction formation in other species, although its regulators and targets important for junction formation are incompletely understood.
We are using C. elegans as a model to understand how aPKC regulates epithelial cell junctions. Using a targeted protein degradation strategy, we first demonstrated that C. elegans PKC-3/aPKC, like PAR-6, is required for junction maturation. A hypomorphic temperature-sensitive allele of PKC-3 causes junction breaks in the spermatheca, a mechanically active organ, leading to sterility. We used this allele in a suppressor screen to find genes that may function with PKC-3 in regulating junctions. We identified multiple intragenic mutations in the PKC-3 gene, including an unexpected stop-to-stop mutation, demonstrating the specificity of the screen. Extragenic suppressors included an allele of the lethal giant larvae gene lgl-1, which antagonizes aPKC in other systems but was not known previously to function in C. elegans epithelia.
Finally, we identified two alleles of a previously uncharacterized gene, sups-1, which encodes an extracellular protein expressed in epidermal cells that localizes to the apical ECM. Our findings define new genes involved in epithelial junction formation in C. elegans and indicate the surprising role that stop codon choice can have on gene activity.
Doctoral Thesis Defenses 2018–19
Below are abstracts defended during 2018–19.
Chelsea Maniscalco—Nance Lab
Title: “Investigating the Remodeling of C. elegans Primordial Germ Cells by Cell Autonomous Contractile Ring-Dependent Cytoplasmic Partitioning”
Date: April 18, 2019
Abstract: Cells such as spermatocytes and erythroblasts form large compartments that are used to eliminate a subset of cellular components. The cellular mechanisms used to create and eliminate such compartments are poorly understood, and it remains unclear whether the varied cell types that dispose of contents in this way employ similar or divergent mechanisms to do so. We previously showed that C. elegans primordial germ cells (PGCs) change shape autonomously to form large organelle-rich lobes (Abdu et al., 2016). PGC lobes are cannibalized by neighboring cells, shrinking PGC volume and eliminating most of their mitochondria. Using time-lapse fluorescence microscopy, we observed that PGC lobes form when the nucleus migrates to one side of the cell and the cell constricts at its equator into two halves—the cell body (containing the nucleus and minimal cytoplasm) and the lobe (enriched in mitochondria).
Because lobe formation resembles an incomplete cytokinesis, we tested whether lobe formation occurs through the assembly and constriction of a contractile ring, even though PGCs are arrested in interphase. Contractile ring components in dividing cells include actin filaments, the motor protein non-muscle myosin, the cross-linking protein anillin, and membrane-associated septin filaments. By expressing fluorescently tagged non-muscle myosin specifically in the PGCs and co-staining with antibodies recognizing other contractile ring components, we observed that a ring containing non-muscle myosin, anillin, and septin localizes at the lobe neck.
To test if ring closure is required for lobe formation, we used temperature-sensitive mutations to acutely block the function of two genes essential for contractile ring formation—myosin II (nmy-2 in C. elegans) and the F-actin nucleator formin (cyk-1 in C. elegans). Inactivation of either component blocked lobe formation. Given that both formin and myosin are found in the PGC lobe contractile ring and are required for lobe formation, the C. elegans RhoA homolog RHO-1 is a likely activator of the lobe neck contractile ring. By using a conditional allele of ect-2, the RhoGEF required to activate RHO-1, we observed that inactivation of ect-2 blocked lobe formation. In dividing cells, ECT-2 is activated by the centralspindlin complex. By contrast, acute inactivation of components of the centralspindlin complex, MgcRacGAP (cyk-4 in C. elegans) and MKLP1 (zen-4 in C. elegans), which localizes to the mitotic spindle in dividing cells and dictates the position of the contractile ring, had no effect on PGC lobe formation. Our findings suggest that an interphase contractile ring, which is assembled and positioned independently of the centralspindlin complex, constricts to partition the PGC cell body from its lobe, allowing PGCs to remodel their size and contents in a spindle-independent manner.
Alice Fok—Ringstad Lab
Title: “Curiouser and Curiouser: Noncanocial Modulators of Bioaminergic and Mechanosensory Systems”
Date: April 15, 2019
Abstract: Cell-intrinsic mechanisms that regulate the activity of dopamine (DA) neurons and their targets are not well understood, and might be important targets for the treatment of neurological and psychiatric disorders caused by dysfunction of brain DA systems. C. elegans offers the opportunity to use behavior-based approaches to discover novel regulators of DA signaling. C. elegans contains eight DA neurons that mediate stereotyped behaviors and are accessible to molecular, genetic, and physiological analysis.
First, I describe experiments that integrate transcriptomic analysis of dopamine neurons and a machine-vision-based behavioral screen to identify genes required for the function of dopamine neurons. Through this approach, I discovered the K2P family K+ channel, TWK-2, which regulates how DA neurons are activated by appetitive food stimuli. TWK-2 channels function in opposition to TRP-4 channels, which are transient receptor potential (TRP) channels required for activation of dopamine neurons. Loss of TWK-2 channels restores physiological function to trp-4 mutant dopamine neurons and also restores a food-response behavior to trp-4 mutants. These data reveal a critical role for a K2P channel in DA neurons and suggest that these ion channels could be therapeutic targets for the treatment of psychiatric and neurological disorders caused by dysfunction of DA systems.
Second, I describe a functional screen for mammalian homologs of C. elegans DA receptors that are ligand-gated ion channels. In vertebrates, DA is thought to signal exclusively through G protein-coupled receptors (GPCRs). The C. elegans nervous system expresses homologs of these GPCRs and also expresses DA-activated Cl- channels that are related to vertebrate GABA and glycine receptors. I hypothesized that there might be vertebrate homologs of these DA-gated channels and performed a functional screen for such homologs. My preliminary results provide support for the hypothesis that DA-gated ion channels can be formed by some assembly of subunits currently annotated as GABAA and glycine receptor subunits. This in turn suggests that DA might directly evoke fast inhibitory signals in mammalian neurons expressing these receptors. Together, my studies reveal novel mechanisms of dopamine signaling that might be conserved between nematodes and vertebrates.
Tugba Colak Champollion—Knaut Lab
Title: “Cadherin-Mediated Cell Coupling Coordinates Chemokine Sensing Across Collectively Migrating Cells”
Date: March 18, 2019
Abstract: Coordinated migration of cell collectives is essential in development, homeostasis, and disease. Cells moving in groups adhere to each other while responding to attractant cues. In many cases, leader cells guide follower cells. However, whether follower cells also contribute to group migration is less clear. To understand how attractant sensing and cell-cell adhesion in leader and follower cells contribute to group migration, I used the zebrafish lateral line primordium (primordium) as a model. The primordium is a collective of about 100 to 150 cells that express the chemokine receptor cxcr4 and migrate in response to a self-generated Sdf1 chemokine gradient. This gradient is linear and stretches across the primordium, suggesting that each cell in the group perceives directional information and could contribute to group movement.
Consistent with this idea, I find that cxcr4 mutant primordia migrate faster as I increase the number of Sdf1 sensing wild-type cells. Although wild-type cells need to be present in the front to restore migration of the cxcr4 mutant primordia, the increase in speed correlates only with the overall wild-type cell number. Also, the ability of wild-type cells to restore migration of cxcr4 mutant primordia suggests that these cells adhere to each other, in such a way that wild-type cells can pull their cxcr4 mutant neighbors along. Likely candidates for cell-cell adhesion in the primordium are cadherins. Consistent with previous reports, I find that E- and N-cadherin are expressed in the primordium. Moreover, I find that these cadherins are required for retaining cells in the primordium and, intriguingly, are essential in wild-type cells to pull cxcr4 mutant primordia along.
Furthermore, analysis of cell dynamics in chimeric and protein-depleted primordia shows that chemokine-sensing and cadherin-mediated cell-cell adhesion contribute jointly to the directional migration of the primordium. Together, these observations suggest that each cell in the primordium extracts directional information from the Sdf1 gradient, contributes to the directional movement of the group, and is coupled to its neighbors through cadherins to ensure efficient migration.
Zharko Daniloski—Smith Lab
Title: “Unique Mechanisms of Sister Chromatid Cohesion at Human Repetitive Sequences”
Date: December 22, 2018
Abstract: Formation of mitotic chromosomes is a complex multi-step process that requires an extensive reorganization of chromatin fibers. Mitotic chromosome organization starts the moment DNA is replicated in S-phase, when the cohesin rings provide sister chromatid cohesion and initiate linear chromatin looping. A second layer of chromosome organization is provided by condensin, which exists as two complexes in humans: condensin I and II. Nuclear condensin II works with topoisomerase II to remove DNA entanglements that result from DNA replication and is required for axial compaction of chromosomes. Following nuclear envelope breakdown, cytoplasmic condensin I gains access to chromosomes and initiates lateral compaction. Together cohesin, condensins, and topo II are required for the formation of individualized mitotic chromosomes and allow proper chromatid segregation in anaphase.
Here I show that human repetitive sequences, telomeres and ribosomal DNA (rDNA), have unique mechanisms of cohesion, resolution, and segregation. Cohesin is comprised of a tripartite ring and a fourth peripheral subunit, which in vertebrates exists in two isoforms: SA1 and SA2. SA1 is required for telomere cohesion, while SA2 is required for centromere cohesion. I showed that SA1 has an AT-hook motif required to directly bind DNA and promote telomere cohesion, independent of the cohesin ring. For SA2, I investigated the observation that STAG2 (the gene encoding SA2) was frequently mutationally inactivated in cancer. I showed that this inactivation resulted in loss of centromere cohesion with a concomitant increase in arm and telomere cohesion that led to reduced telomere shortening and increased lifespan in normal human cells.
Finally, I found that the human rDNA arrays resolved and segregated after the rest of the genome in anaphase, dependent on the PARP tankyrase 1, condensin II, and topo II. Defective rDNA resolution led to increased DNA damage, micronuclei formation, and aneuploidy. Altogether, my thesis uncovers unique mechanisms of cohesion at repetitive sequences that are required to maintain genome stability.
Charles Ng—Littman Lab
Title: “Silencing CD4 in CD8+ T Cells: Insights into Lineage Commitment and Heritable Gene Repression”
Date: September 20, 2018
Abstract: During development, the transcriptional repression of alternative lineage genes is critical for cell fate commitment. CD8+ cytotoxic T cells and CD4+ helper/regulatory T cells develop from a bipotent CD8+CD4+ DP cell in the thymus in a process referred to as lineage commitment. Over several decades of study, many of the transcription factors required for the development of these T cell lineages have been identified. However, the underlying mechanisms governing their differentiation remain unknown.
To expand our understanding of lineage commitment we have focused on studying the regulation of the Cd4 gene, as it shows a near invariant correspondence with T cell lineage. Cd4 silencing in the cytotoxic lineage is initiated by a negative regulatory element in cis at the Cd4 locus, but once established, becomes independent of this DNA sequence suggesting epigenetic mechanisms subsequently maintain its repression.
To study the maintenance of silent gene states, we investigated how the Cd4 gene is stably repressed in cytotoxic CD8+ T cells. Through CRISPR and shRNA screening, we identified the histone chaperone CAF-1 as a critical component in the silencing machinery. The large subunit of CAF-1, Chaf1a, associates with the histone deacetylases Hdac1/2 and the histone demethylase Lsd1, enzymes that we found to also be required for Cd4 silencing.
In the absence of the DNA methyltransferase Dnmt3a, but not the replication-coupled maintenance DNA methyltransferase Dnmt1, there was a markedly increased sensitivity to Cd4 derepression mediated by CAF-1 deficiency. In contrast to Dnmt1, Dnmt3a did not grossly affect levels of DNA methylation at the Cd4 locus. Instead, Dnmt3a deficiency sensitized CD8+ T cells to Cd4 derepression mediated by histone modifying factors, including the enzymes associated with CAF-1.
Thus, we have uncovered cooperative functions among the DNA methyltransferases and CAF-1 in the heritable silencing of the Cd4 gene in CD8+ T cells, providing new insight into the regulation of cell identity.
Annabelle Suisse—Treisman Lab
Title: “Calcium Homeostasis and Proteasomal Degradation Regulate Transcription Factor Availability Downstream of Conserved Signaling Pathways”
Date: September 17, 2018
Abstract: A handful of conserved signaling pathways govern cell proliferation, fate determination, and behavior throughout development and during adult life. Proper integration of signals received by the cell is crucial to prevent developmental defects and diseases such as cancer. The most downstream effectors of these signaling pathways are transcription factors. In the Wingless (Wg) pathway, Armadillo (Arm) is protected from proteasomal degradation and activates target gene transcription when the pathway is active. In the Epidermal Growth Factor Receptor (EGFR) pathway, the transcriptional repressor Capicua (Cic) is disabled upon phosphorylation by active mitogen-activated protein kinase (MAPK).
We found that mutations in the endoplasmic reticulum (ER) calcium pump SERCA disrupt Wg signaling by sequestering Arm away from the signaling pool. SERCA affects calcium (Ca2+) homeostasis in the lumen of the ER, which is required for correct processing and trafficking of transmembrane proteins. In SERCA mutants, Arm remains bound to the adherens junction protein E-cadherin, which is retained in the endoplasmic reticulum when Ca2+ levels are reduced.
Using hypomorphic and null SERCA alleles in combination with loss of the plasma membrane calcium channel Orai allowed us to define three distinct thresholds of ER Ca2+, which differentially affect Wg signaling and two other pathways: Notch and Hippo.
Second, we found that mutations in the COP9 signalosome subunit 1b (CSN1b) cause ectopic expression of EGFR target genes. The COP9 signalosome (CNS) removes Nedd8 peptides from the Cullin subunits of ubiquitin ligase complexes, reducing their activity.
Analysis of CSN1b and mutations affecting other CSN subunits revealed that the CSN complex protects Cic from EGFR pathway-dependent ubiquitination by a Cullin 1/SKP1-related A/Archipelago E3 ligase and subsequent proteasomal degradation. The CSN1b subunit also maintains basal Capicua levels by protecting it from a separate mechanism of degradation that is independent of EGFR signaling.
Additionally, we used an uncleavable form of Nedd8 to provide in vivo evidence for deneddylase-independent inhibition of Cullin activity by the CSN.
Together, these studies highlight the importance of protein synthesis and degradation in regulating the abundance of transcription factors downstream of signaling pathways.
Jessica Douthit—Treisman Lab
Title: “Lost and Found and Plexin A Contribute to Visual Circuit Development in Drosophila”
Date: August 20, 2018
Abstract: The Drosophila visual system is an excellent model system for studying the basic mechanisms of axon pathfinding and neural circuit formation. The terminals of R7 and R8 photoreceptors, responsible for color vision, are segregated into distinct target layers of the medulla, a central visual processing center in the brain. Ultraviolet-sensitive R7 axons terminate in the M6 layer, while blue/green-sensitive R8 axons target the M3 layer.
We have found that lost and found (loaf), which encodes a CUB-LDL protein, is required in R7 photoreceptors for their normal targeting to the M6 layer. Although flies homozygous mutant for loaf show normal R7 projections, loaf mutant R7 axons in a wild-type background or R7 photoreceptors expressing loaf RNAi prematurely terminate in the same layer as R8 axons. We believe that R7 cells may use Loaf to compete with another cell type in the medulla for a ligand expressed by a common target cell.
To search for such a ligand, we performed an RNAi screen for genes required in medulla neurons for R7 targeting. We found that knocking down plexin A (plexA) in medulla neurons caused R7s to mistarget to the M3 layer and fail to expand their axon terminals. Shortened R7 projections were also observed in plexA mutants.
We have used RNAi and somatic CRISPR to show that tangential fibers that cross the medulla in the M7 layer are the most likely source of PlexA. The effect of plexA on R7 targeting may be indirect, as plexA mutations affect the overall structure of the medulla; markers of different medulla layers are still present but show reduced separation. The Sema domain of PlexA, which engages Semaphoring binding partners, is required for this function.
However, we found that Semaphorin-1a and Semaphorin-1b are not required in photoreceptors for R7 targeting. Defects in plexA mutants are more severe than those observed when loaf levels are altered. Misexpression of PlexA on photoreceptor axons causes them to adhere to one another, and this effect does not require loaf in the axons, arguing that Loaf is not a PlexA receptor.
Overall, this work identified Loaf and PlexA as cell surface molecules that provide specific instructions for the development of the Drosophila visual system, and may provide insights into how similar molecules behave in higher organisms.
Natasha Koppel—Burden Lab
Title: “Characterization of Vezatin, an Adherens Junction Protein, Required for both Myogenesis and the Maintenance and Maturation of Neuromuscular Synapses”
Date: August 7, 2018
Abstract: Understanding the mechanisms responsible for forming and maintaining neuromuscular synapses remains a paradigm not only for unraveling how synapses are built and stabilized but also for understanding the principles and mechanisms that underlie the formation of other specialized cell-cell junctions.
Key genes, such as Agrin, Lrp4, MuSK, and Rapsyn are required both for the formation and postnatal maturation of mammalian neuromuscular synapses, including the initial clustering, anchoring, and subsequent rearrangement of acetylcholine receptors (AChRs) in the postsynaptic muscle membrane.
Here, we show that vezatin, a widely expressed integral membrane protein and component of adherens junctions, is required for two processes in skeletal muscle. First, vezatin inactivation in myoblasts leads to a reduction in the number of myoblasts and myotubes, causing neonatal lethality. Second, we demonstrate that vezatin binds directly to AChRs and is required for the maturation but not formation of neuromuscular synapses.
In the absence of skeletal muscle vezatin, neuromuscular synapses fail to fully transition from a plaque-like shape to a complex, branched shape; postjunctional folds form aberrantly; and the density of postsynaptic AChRs is reduced. Agrin stimulates AChR clustering in vezatin-deficient myotubes, but these AChR clusters are labile and disassemble following Agrin withdrawal, suggesting a role for vezatin in stabilizing AChRs.
Carolyn A. Morrison—Treisman Lab
Title: “A Single Transcription Factor Specifies Multiple Cell Fates in the Drosophila Eye”
Date: July 16, 2018
Abstract: The Drosophila eye consists of photoreceptor neurons and two types of non-neuronal support cells, the lens secreting cone cells and optically insulating pigment cells, all derived from a common pool of progenitors. These different cell types are specified by the combinatorial use of cell type-specific transcription factors and extrinsic signaling cues. This combinatorial code ensures the appropriate activation of distinct gene networks required for terminal differentiation.
Previous reports indicated that glass (GI), a zinc finger transcription factor, acted solely in neurons as a photoreceptor determinant. However, GI is broadly expressed in the eye from the onset of differentiation in both neuronal and non-neuronal cell populations, raising the possibility that GI may contribute to non-neuronal eye cell development.
To determine whether GI could only function in neuronal lineages, as suggested by previous reports, we misexpressed GI in larval neuroblasts and in epithelial cells in the wing disc. We determined that GI was able to activate some target genes in both of these cell types. A comparative RNA sequencing analysis revealed that GI induced overlapping but distinct gene sets in the two tissues, including markers of photoreceptors, cone cells, and pigment cells. Induction of some of these genes was further potentiated by receptor tyrosine kinase (RTK)/Ras signaling, which is reiteratively required throughout eye development.
Moreover, we identified additional transcription factors that synergize with GI to promote the expression of non-neuronal cone and pigment cell markers. Cell type-specific somatic clustered regularly interspaced short palindromic repeats (CRISPR) and rescue experiments demonstrated a cell-autonomous requirement for GI in promoting cone cell and pigment cell development.
Together, these results indicate that GI is autonomously required to promote the development of all the cell types which make up the Drosophila eye.
Doctoral Thesis Defenses 2017–18
Below are abstracts of earlier defenses.
Sarah Cantor—Burden Lab
Title: “Retrograde Signaling at the Neuromuscular Junction: Mechanisms and Therapeutic Approaches to Stimulate Differentiation and Attachment of Motor Nerve Terminals to Muscle”
Date: April 23, 2018
Abstract: Motor neurons and skeletal muscle exchange signals to ensure differentiation of presynaptic and postsynaptic membranes at the neuromuscular synapse. Agrin, provided by motor neurons, binds to lipoprotein receptor–related protein 4 (Lrp4) in muscle, which stimulates association between Lrp4 and muscle-specific kinase (MuSK), leading to MuSK activation. Tyrosine-phosphorylated activated MuSK initiates a pathway that anchors key postsynaptic proteins, including acetylcholine receptors (AChRs) and Lrp4, in the postsynaptic membrane and induces expression of the genes encoding these postsynaptic proteins in subsynaptic nuclei. Once clustered by MuSK signaling, Lrp4 functions as a direct retrograde signal to motor nerve terminals to induce presynaptic differentiation. One aim of my thesis has been to identify and study the motor neuron receptor that binds and responds to Lrp4 to induce presynaptic differentiation.
Here, I describe two family members who are candidate Lrp4 receptors (Lrp4-Rs) and provide evidence that these receptors play an important role in neuromuscular synapse formation. First, I show that Lrp4-expressing cells induce presynaptic differentiation in motor neurons derived from ES cells but not in mutant, ES cell–derived motor neurons deficient in these Lrp4-Rs. Second, I show that at least one of the Lrp4-Rs is concentrated at the neuromuscular synapse. Third, mice deficient in the Lrp4-Rs display defects in synapse formation, suggesting that these Lrp4-Rs respond to Lrp4 to regulate presynaptic differentiation.
In the second half of my thesis work, I describe experiments to test whether boosting retrograde signaling might promote and maintain attachment of motor nerve terminals to muscle in denervating diseases, such as amyotrophic lateral sclerosis (ALS). In ALS, motor nerve terminals detach from muscle and subsequently undergo cell death, together leading to progressive motor weakness and ultimately respiratory paralysis. I show that increasing MuSK activity, using an agonist antibody to MuSK, decreases muscle denervation, improves motor system output, reduces motor neuron cell death, and extends the lifespan of SOD1 G93A mice, an aggressive model for ALS. These results reveal a novel therapeutic strategy for ALS, which targets a key signaling pathway for maintaining attachment of nerve terminals to muscle.
Simon Vidal—Stadtfeld Lab
Title: “Signaling Pathway Modulation Identifies the H3K9 Methyltransferase GLP as a Positive Regulator of Efficient Cellular Reprogramming”
Date: April 10, 2018
Abstract: The reprogramming of somatic cells into induced pluripotent stem cells (iPSCs) provides an elegant system to study mechanisms underlying cell fate change. Due to the low frequency (approximately 1 percent of input cells) and considerable lag phase (more than 12 days) of reprogramming, key aspects of iPSC formation remain poorly understood. My thesis work aimed to utilize the rationale modulation of cellular signaling pathways to develop and characterize a rapid and efficient reprogramming system. I discovered that inhibition of TGFβ together with the activation of Wnt signaling in the presence of ascorbic acid allows the vast majority of murine fibroblasts to acquire pluripotency after one week or reprogramming factor expression. We called this approach three compounds (3C) reprogramming. In contrast to the complex requirements of fibroblasts to reprogram efficiently, hepatic and myeloid blood progenitor cells converted into iSPCs at efficiencies approaching 100 percent upon TGFβ inhibition or canonical Wnt activation, respectively.
Surprisingly, 3C reprogramming of fibroblasts requires the enzymatic activity of the H3K9me1/2 methyltransferase GLP, which under basal conditions constitutes a well-established roadblock of iPSC formation due to its role in silencing of pluripotency-associated gene loci. Genome-wide transcriptional profiling revealed that 3C reprogramming is characterized by the efficient downregulation of mesenchymal genes and an enhanced metabolic reprogramming, two features differentially affected by the loss of GLP activity. Specifically, inhibition of GLP resulted in only partial repression of key mesenchymal transcription factors and an inefficient activation of genes involved in the serine-glycine and one-carbon metabolism. Notably, exogenous S-Adenosyl methionine (SAM), a one-carbon metabolite and cofactor of DNA and histone methyltransferases, partially rescued the reprogramming upon GLP inhibition in 3C conditions. We provide evidence that ascorbic acid in a dose-dependent manner establishes the requirement for GLP activity during enhanced reprogramming.
Altogether, my thesis describes cell type–specific modulation of signaling pathways that can be employed to achieve efficient and synchronous iPSC formation from different somatic cells. Furthermore, my results establish GLP as a context-dependent regulator of cell fate change and provide evidence for a previously unappreciated interaction between H3K9 methylation and metabolic reprogramming during the induction of pluripotency.
Alexandra Pinzaru—Sfeir Lab
Title: “POT1 Dysfunction Causes Telomere Replication Stress and Fuels Tumorigenesis”
Date: March 28, 2018
Abstract: Chromosome ends are protected from being recognized as DNA breaks by a subset of highly specialized proteins that includes protection of telomeres 1 (POT1). Genome sequencing studies have associated a number of POT1 mutations with a variety of cancer types; however, the contribution of these POT1 variants to disease pathogenesis has not been experimentally addressed. My thesis work aimed to comprehensively characterize the functional consequences of two cancer-associated POT1 mutations and to determine whether a dysfunctional POT1 could influence cancer progression. My research focused on the p.F62V and p.K90E substitutions, which affect conserved residues in POT1 and were identified in cutaneous T-cell lymphoma patients. The p.K90E variant was additionally found in a case of chronic lymphocytic leukemia and in a large family afflicted by cancer predisposition, primarily melanoma.
I show that both mutations disrupt normal POT1 function, leading to ATR-dependent DNA damage activation at telomeres, mild chromosome fusions, and telomere elongation. Moreover, my results suggest POT1 plays a role in facilitation replication in the chromosome ends, as the two mutants cause fragility and replication fork stalling at telomeres. My data links these phenotypes to impaired CST (CTC1-STN1-TEN1) function at telomeres. In addition, I find that deletion of murine Pot1a in common lymphoid progenitor cells cooperates with p53 loss to accelerate the onset and severity of T-cell lymphomas. Moreover, the Pot1a−/−Trp53−/− tumors have highly rearranged genomes compared to control Trp53−/− lymphomas and downregulate the ATR signaling pathway in order to limit checkpoint activation.
Lastly, I further investigate how cells carrying POT1 mutations cope with genome instability by conducting a genome-wide, synthetic lethality screen in human cells. I complement my screen with large-scale proteomics to determine potential unknown POT1 interactors and proteins involved in response to telomere replication stress. My preliminary analysis suggests the POT1 mutants genetically interact with factors necessary for DNA replication and repair, and for ensuring faithful chromosome segregation.
Altogether, my thesis characterizes a previously unappreciated mechanism by which defective telomere replication can fuel cancer progression and provides insight into the potential adaptive mechanisms and vulnerabilities of cancer cells carrying POT1 mutations.
Juhee Pae—Lehmann Lab
Title: “Precise Protein Degradation Separates Germline from Soma”
Date: February 26, 2018
Abstract: Germ cells are specialized cells with the potential to produce an entirely new organism. The separation of germline from somatic lineages is one of the first key decisions made during embryonic development in sexually reproducing species and is, therefore, fundamental to reproduction and species preservation. In Drosophila melanogaster, an evolutionarily conserved BTB-domain protein, Germ cell-less (GCL), is essential for the proper formation of the germline precursors. At the onset of my thesis, however, the precise molecular mechanisms by which GCL promotes the establishment of the germline–soma dichotomy were uncertain. My thesis research demonstrates that GCL is a substrate recognition subunit for the Cullin3-RING ubiquitin ligase complex (CRL3GCL) and acts as a molecular switch that turns off a somatic lineage pathway specifically in the germline through ubiquitin-mediated protein destruction.
Combining genetic suppression and biochemical experiments, I show that CRL3GCL promotes ubiquitylation and degradation of Torso, a receptor tyrosine kinase (RTK) and a major determinant of somatic cell fate. Disruption of CRL3GCL results in inappropriate Torso accumulation, thereby activating the somatic program at the expense of the germline. Intriguingly, unlike the classical paradigm for RTK degradation, CRL3GCL-mediated degradation of Torso RTK does not depend upon receptor activation. It is instead prompted by a cell cycle–dependent change in GCL localization and subsequent ubiquitylation of Torso specifically during mitosis, thus revealing a novel mode of RTK regulation. Moreover, utilizing genetic and optogenetic tools, I show that Torso antagonizes germline formation independently of its canonical Ras/Raf/MAPK signaling cascade, which is required for transcription of somatic signatures. Further investigation of these previously uncharacterized components will therefore shed light on how the bifurcation of a signaling pathway couples cell formation with fate specification.
Altogether, these findings illustrate an exquisitely coordinated mechanism that is essential to define cell lineage at the single-cell level and to promote the germline–soma dichotomy required for species propagation through generations. The mechanistic details described in my thesis also have broader implications for our understanding of the cellular events, such as cell signaling and cytoskeletal organization, that coordinate the development of different cell types.
Kara Zang—Ringstad Lab
Title: “Modulating the Modulator: Mechanisms That Regulate Serotonin Neurons In Vivo”
Date: January 15, 2018
Abstract: Serotonin is a conserved neuromodulator that is involved in numerous physiological processes and functional states, including feeding, thermoregulation, cognition, and affect. Although much is known about how serotonin acts on its cellular targets, how its release is regulated in vivo remains poorly understood. In the nematode C. elegans, female reproductive behavior depends on a pair of serotoninergic motor neurons, which are directly modulated by inhibitory neuropeptides.
Here, I report the isolation of mutants in which inhibitory neuropeptides fail to properly modulate serotonin neurons and the behavior they mediate. One of the corresponding mutations affects the T-type calcium channel CCA-1 and symmetrically retunes the voltage dependencies of its activation and inactivation toward more hyperpolarized potentials. This shift in voltage dependency strongly and specifically bypasses the effects of peptidergic inhibition on serotonin neurons. The other corresponding mutation isolated from this screen likely affects the two-pore potassium channel TWK-17. The putative TWK-17 mutation restores activity to serotonin neurons that have been shut down due to excess peptidergic inhibition. My results indicate that manipulations of these two types of ion channels can restore function to hypo-active neurons and may demonstrate conserved mechanisms governing serotonin release.
Yusuff Abdu—Nance Lab
Title: “Investigating Developmentally Programmed Germ Cell Remodeling by Endodermal Cell Cannibalism”
Date: January 8, 2018
Abstract: During development, primordial germ cells (PGCs) form early and undergo several specialized molecular and cellular remodeling events to establish and maintain the germ cell program. A conserved behavior of PGCs is to intimately associate with endodermal cells. Yet, the significance of this germ cell–endoderm association remains largely unexplored. Midway through the embryogenesis of Caenorhabditis elegans, PGCs form large protrusions called lobes that become embedded inside the adjacent endodermal cells. The fate of the lobes and the significance of PGC–endoderm interaction has remained a 30-year-old mystery.
Using live fluorescent imaging, I show that PGC lobes are actively removed and degraded by endodermal cells, dramatically reducing PGC size and mitochondrial content. I demonstrate that endodermal cells do not scavenge lobes PGCs shed, but rather actively cannibalize lobes from the PGC cell body through a developmentally regulated mechanism. Also, I find that endodermal CED-10/Rac1-induced actin, DYN-1/dynamin, and LST-4/SNX9 transiently surround lobe necks and are required for lobe scission. My results suggest that scission occurs through a mechanism analogous to mitochondrial fission in animal cells where a constriction machinery in the mitochondrial outer membrane promotes scission in both outer and inner membranes. Interestingly, I find that PGC mitochondria are enriched in potentially damaging reactive oxygen species (ROS) before their elimination. I hypothesize that lobe cannibalism protects PGCs by reducing mitochondrial-produced ROS within PGCs.
These findings reveal an unexpected role for endoderm in altering the contents of embryonic PGCs and define a form of developmentally programmed cell remodeling involving intercellular cannibalism. Active roles for engulfing cells have been proposed in several neuronal remodeling events, suggesting that intercellular cannibalism may be a more widespread method used to shape cells.
Yolande Grobler—Lehmann Lab
Title: “Genome-Wide Analysis of Wolbachia–Host Interactions”
Date: December 18, 2017
Abstract: Insects are common vectors for devastating human viruses such as Zika and dengue. A novel strategy for preventing the transmission of vector-borne viruses exploits the bacterium Wolbachia. Wolbachia is suspected to infect 70 percent of insect species and provides host insects with resistance to many viruses. Therefore, the replacement of insect populations with those infected with Wolbachia can be used to drive down vector-borne virus transmission. To ensure the success of this replacement strategy, the interactions between Wolbachia and insect hosts need to be understood. For my thesis, I focused on two aspects of Wolbachia–host interaction: which host cellular compartments harbor Wolbachia and how host systems affect Wolbachia load dynamics.
To address these questions, I used the Wolbachia/Drosophila model that provides a genetically tractable system for studying host–pathogen interactions. First, using serial focused ion beam electron microscopy, I found the Wolbachia intercellular niche to be of host endoplasmic reticulum origin. Second, I used a Wolbachia-infected Drosophila cell line JW18 to perform an unbiased Drosophila whole-genome RNAi screen taking advantage of a novel high-throughput fluorescence in situ hybridization (FISH) assay to detect changes in Wolbachia levels. 1,117 genes altered Wolbachia when silenced by RNAi, of which 329 genes increased and 788 genes decreased the level of Wolbachia. Validation of hits included in-depth secondary screening using in vitro and in vivo RNAi, Drosophila mutants, and qPCR. A diverse set of host gene networks were identified to regulate Wolbachia levels, including transcription, epigenetic modification, protein anabolism and catabolism, translation, and cell cycle. I focused on an unexpected role of host proteasome and host translation factors such as the ribosome and translation initiation factors in suppressing Wolbachia levels both in vitro and in vivo.
Preliminary results suggest that these interactions may have implications in the Wolbachia-mediated antiviral response in hosts. This screen provides the most comprehensive analysis of Wolbachia–host interactions to date and can serve as the basis for future research in understanding the molecular mechanisms that govern this unique symbiotic relationship.
Luis A. Martínez-Velázquez—Ringstad Lab
Title: “Molecular Genetics of Trafficking to the Sensory Cilium”
Date: September 8, 2017
Abstract: The primary cilium is a microtubule-based organelle specialized for cell signaling. Mutations that affect the generation and maintenance of the primary cilium are the root cause of a family of disorders that affect multiple organ systems and are collectively referred to as ciliopathies. Defects caused by ciliopathies include situs inversus, obesity, intellectual disability, polycystic kidney disease, and sensory dysfunction such as anosmia and blindness. The sensory defects caused by ciliopathies highlight the important role of the primary cilium in sensory neurons, which use cilia to organize factors involved in sensory transduction. Despite the importance of cilia in sensory transduction, how proteins are targeted to these specialized membrane domains remains poorly understood. Sensory neurons of the nematode C. elegans offer an excellent model to use genetic analysis to determine mechanisms required for the generation, maintenance, and function of sensory cilia.
Through studies of the chemosensory BAG neurons of C. elegans, I have identified the nematode homologue of a human retinal dystrophy gene of unknown function, RD3. I show that this gene encodes a Golgi-associated protein required for efficient trafficking of receptor-type guanylate cyclase to the sensory cilium. Furthermore, I use this model to identify mutations that restore cyclase trafficking to RD3 mutants. Suppressor mutations target key components of the retromer complex, which mediates recycling of cargo from endosomes to the Golgi.
My data show that in sensory neurons a critical balance exists between the rates of anterograde and retrograde trafficking of cargo destined for the sensory cilium and that this balance requires a molecular specialization at an early stage of the secretory pathway. Finally, I show that this experimental model can be used to identify additional factors that likely promote the transport of cargo containing vesicles to the base of the sensory cilium and fusion to the ciliary membrane.
Aaron Phillips—Sfeir Lab
Title: “Single-Molecule Analysis of mtDNA Replication Uncovers the Basis of Common Deletion”
Date: July 7, 2017
Abstract: Mutations in mitochondrial DNA (mtDNA) lead to muscular and neurological diseases and are linked to aging. The most frequent aberrancy is the “common deletion” that involves a 4,977-bp region flanked by 13-bp repeats. To investigate the basis of this deletion, we developed a single-molecule mtDNA combing method. The analysis of replicating mtDNA molecules provided in vivo evidence in support of the asymmetric mode of replication. Furthermore, we observed frequent fork stalling at the junction of the common deletion, suggesting that impaired replication triggers the formation of this toxic lesion. In parallel experiments, we employed mito-TALENs to induce breaks in distinct loci of the mitochondrial genome and found that breaks adjacent to the 5’ repeat trigger the common deletion. Interestingly, this process was mediated by the mitochondrial replisome independent of canonical DNA double-stranded break repair. Altogether, our data underscore a unique replication-dependent repair pathway that leads to the mitochondrial common deletion.
Vanguel Trapkov—Stadtfeld Lab
Title: “Transgenic Systems to Study the Transcriptional Regulation of In Vitro Hematopoietic Progenitor Cell Emergence”
Date: June 15, 2017
Abstract: Blood is maintained by hematopoietic stem and progenitor cells (HSPCs), which constantly replace mature blood cell types lost to injury or turnover and which are under the tight control of signaling pathways and transcriptional regulators. Unraveling this interplay is instrumental for our understanding of hematological malignancies and the generation of HSPCs in vitro.
Our work focused on studying transcriptional requirements of specifying HSPCs from pluripotent as well as non-hematopoietic cells in vitro, with the goal of understanding the process of hematopoietic commitment during mammalian hematopoietic development, but also addressing the critical clinical need of the shortage of patient-matched cells for bone marrow transplantation. Specifically, we aimed to generate transgenic systems to allow (a) understanding the role of the critical transcription factor (TF) RUNX1 in driving the developmental conversion of hemogenic endothelium (HE) into HSPCs; (b) the identification of co-regulators of RUNX1 function in this process; and (c) the study of two hematopoietic TF triads—RUNX1-SCL-GATA2 and FOS-GFI1BGATA2—in their ability to initiate hematopoiesis.
Our studies revealed an important dosage requirement of RUNX1 for the kinetics of HSPC emergence and the functional properties of generated blood progenitors. In addition, we identified a set of HE-associated transcriptional regulators that fail to be appropriately silenced in HSPCs in vitro and whose importance for HSPC quality we have evaluated. Finally, we show that FGG can expand pre-existing macrophages in skin biopsies and that both triads support HSPC specification during directed differentiation. Overall, our studies demonstrate the importance of precise TF regulation during the emergence of HSPCs and open new avenues for the studies of combinatorial TF interactions during in vitro blood generation.
Evgenia Korol—Tahiliani Lab
Title: “In Search of Novel TET2 Functions”
Date: June 15, 2017
Abstract: Organisms use DNA as a blueprint for encoding instructions to our cells, maintaining their life cycle, and supporting the propagation of future generations. The DNA code is tightly regulated—depending on the cell type and environmental conditions, certain regions of this genetic blueprint are more or less accessible to the cell. One form of regulation is methylation of the fifth carbon position of the cytosine base to form 5-methylcytosine (5mC) within our genome. Cytosine methylation affects all aspects of cellular processes and plays an important role in areas ranging from X-chromosome inactivation to prevention of disease development. The study of this mechanism promises to give us the ability to better control cellular processes and prevent disease. The Ten-Eleven Translocation proteins (TET1, TET2, and TET3) are able to oxidize the methyl group on 5mC.
In order to learn about the function of these TET modifications, we performed the tandem affinity purification (TAP) experiment to identify proteins that associate with TET2. By identifying proteins that interact with TET2, we sought to learn and understand the implications of TET enzymatic activity in cells. We identified P54NRB/Nono, PSPC1, and SFPQ as associated partners of TET2. Finally, we showed that the interaction between TET2 and P54NRB/Nono could be occurring at a subset of bivalent genes in mESC, suggesting their potential role in the regulation or maintenance of the bivalency of these genes during the early development of an organism.
John Wang—Knaut Lab
Title: “Local Changes in the Extracellular Matrix Facilitate FGF Diffusion for Focal Signaling”
Date: June 6, 2017
Abstract: Cells communicate with each other using soluble signaling molecules (ligands) to coordinate cellular patterning, growth, and movement during embryonic development. Many ligands interact with components of the extracellular matrix (ECM), which alter their mobility, stability, and activity. However, it remains mostly unclear how this regulation occurs at the single-molecule level. The fibroblast growth factor (FGF) ligand Fgf10a coordinates sensory organ formation in the zebrafish posterior lateral line primordium.
Using fluorescence correlation spectroscopy, I analyzed the diffusion behavior of individual Fgf10a molecules and the effect of the ECM protein Kal-1 (Anosmin-1), mutations in which cause Kallmann syndrome in humans, on this behavior. I find that most Fgf10a move rapidly through the ECM, resulting in high overall Fgf10a diffusivity, with a second minor population of slowly diffusing Fgf10a that is presumably bound to ECM components such as heparin sulfate. Kal-1 enhances Fgf10a diffusivity by shifting the equilibrium between the fast and slow populations of Fgf10a to favor the fast population. This allows Fgf10a to diffuse from the front of the primordium to the rear, where it enriches in microlumens to focally concentrate FGF signaling.
These observations reveal a thus far unappreciated mechanism whereby an extracellular modulator of diffusion locally enhances ligand diffusivity in the ECM to concentrate the ligand for focal signaling. The use of such diffusion modulators may be a general mechanism for adjusting signaling in multicellular organisms.