Investigations Into Roles for Endocytosis in LIN-12/Notch

Investigations into roles for endocytosis in LIN-12/Notch

signaling and its regulation

Jessica Chan

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy under the Executive Committee of the Graduate School of Arts and Sciences

COLUMBIA UNIVERSITY

2020

Jessica Chan

Abstract

The LIN-12/Notch signaling pathway is highly conserved in all animals, and is crucial for proper development. It is a key pathway in specifying cell fate in many cellular contexts, and dysregulation of the pathway can have deleterious consequences. Therefore, understanding how

LIN-12/Notch signaling is regulated in different contexts has been a main area of interest in the field. Previous studies in different model organisms have identified many modes of regulation of the signaling pathway, one of which is endocytosis of the ligand and receptor. Here, I further investigated the role of endocytosis in LIN-12/Notch signaling in multiple developmental contexts in Caenorhabditis elegans.

Work in Drosophila and vertebrates had previously established that ligand-mediated activation of Notch requires ubiquitination of the intracellular domain of the transmembrane ligand and the activity of the endocytic adaptor Epsin in the signaling cell. The consensus in the field is that Epsin-mediated endocytosis of mono-ubiquitinated ligand generates a pulling force that exposes a cleavage site in Notch for an ADAM protease, a critical step in signal transduction. In contrast, in this thesis, I examined two different transmembrane ligands in several different cell contexts and found that activation of LIN-12/Notch and the paralogous

GLP-1/Notch in C. elegans does not require either Epsin-mediated endocytosis or ubiquitination of the intracellular domain of the ligand. Results obtained by a collaborator indicate that C. elegans ligand and receptor interactions are tuned to a lower force threshold than are Drosophila ligand and receptor interactions, potentially accounting for these differences.

I also looked at the role of endocytosis in regulating LIN-12 signaling in the context of vulval development. The cell fate pattern of six vulval precursor cells (VPCs) is mediated by

EGFR and LIN-12/Notch signaling. Previous work using multicopy transgenes in fixed

specimens indicated that LIN-12 is post-translationally downregulated via endocytosis in response to EGFR activation in the VPC named P6.p, an event that appeared essential for ligands to activate LIN-12/Notch in neighboring VPCs. In this thesis, I manipulate the endogenous lin-

12 gene and examine live specimens to show that LIN-12 appears to be regulated transcriptionally in P6.p and evidence that there may be additional potential endocytic motifs that may regulate LIN-12 in this context.

Table of Contents

List of Tables and Figures...... iii

Acknowledgments...... v

Chapter 1: Introduction ...... 1

1.1 Overview of Notch signaling in C. elegans ...... 2

1.2 Overview of Notch signal transduction ...... 6

1.3 Notch activation by DSL ...... 10

1.4 Role of secreted DSL ...... 13

1.5 Known roles of epn-1/Epsin in C. elegans ...... 14

Chapter 1 Figures ...... 17

Chapter 2: Results ...... 27

Summary ...... 28

Results ...... 29

Chapter 2 Figures ...... 37

Chapter 3: Discussion ...... 48

Chapter 3 Figures ...... 58

Chapter 4 Materials and Methods ...... 61

References ...... 74

Appendix ...... 87

Appendix A: Introduction ...... 88

Introduction Figures ...... 97

Appendix B: Results ...... 102

Results Figures ...... 111

Appendix C: Discussion ...... 126

Appendix D: Materials and Methods ...... 132

List of Tables and Figures

Chapter 1. Figures……………………………………………………………….……….………18

Figure 1. Notch Signaling in C. elegans embryogenesis…….………………………….….……18

Figure 2. Notch Signaling in C. elegans germline development.…………………….……….…19

Figure 3. Notch Signaling in C. elegans AC/VU decision.………………………….….……….20

Figure 4. Notch Signaling in C. elegans VPC specification ………………….…….….………..21

Figure 5. Overview of Notch signaling cascade…………...………………….…….….………..22

Figure 6. DSL in C. elegans, Drosophila and vertebrates…………...…..…….…….…………..23

Figure 7. Domain organization of C. elegans, Drosophila and human Notch………..…………24

Figure 8. Epsin domain structure and function………….……..……………………..………….25

Figure 9. An In vivo force sensor in Drosophila…………………………….………..…………26

Chapter 2. Figures……………………………………………………………….……………….39

Figure 1. epn-1 is not required for GLP-1 signaling at 4 cell stage ……………………………..39

Figure 2. epn-1 is not required for LAG-2 function in specifying excretory cell ……....………40

Figure 3. Ubiquitination is not required for LAG-2 function in embryogenesis and AC/VU decision…………………………………………………………………………………………..41

Figure 4. Membrane anchor is required for ligand function……………………………………..42

Supplemental figure 1. epn-1::zf1::GFP expression in L3……………………………………...43

Supplemental figure 2. A difference in LAG-2 intracellular domain requirement in AC/VU decision……………………………………………………………………………………...…...44

Chapter 3. Figures………………………………………………………………………………..59

Figure 1. Predicted DSL in Bilateria and non-Bilateria………………………………………….59

Figure 2. Alignment of NRR sequence from human, Drosophila, and nematode Notches……...60

iii

Appendix Figures……………………………………………………………….….….……….98

Introduction Figure 1. Vulval Precursor Cell specification ……………………………………..98

Introduction Figure 2. LIN-12 expression in VPCs…………….……………………..…...... 99

Introduction Figure 3. LIN-12 internalization by Downregulation Targeting Sequence

(DTS)………………………………………………………………………..………………….100

Introduction Figure 4. LIN-12 downregulation mediated by alx-1 and wwp-1…..… .……..….101

Results Figure 1. VPC-specific RNAi screen…………………..……………………..….…….112

Results Figure 2. Single copy transgenes of lin-12 and lin-12 mutants… ...………………...…113

Results Figure 3. Summary of lin-12(DTS)::GFP multicopy transgene expression…..…..…...114

Results Figure 4. Endogenous expression of lin-12::GFP and lin-12Δ(DTS)::GFP ……...... 115

Results Figure 5. Endogenous lin-12 transcriptional reporter.....……………………..………..116

Acknowledgments

I am writing this amid the COVID-19 pandemic, serving as a stark reminder of the privileges I have and the many things that I am grateful for throughout this journey. First and foremost, I am deeply grateful for my advisor, Dr. Iva Greenwald, for her guidance throughout my time in the lab. There were many times when I encountered obstacles during my research, as well as in personal life, and she has always been patient and provided valuable advice. Her dedication to both teaching and life-long learning, as well as the creativity and rigor she applies to science has been beyond inspiring and certainly worth emulating no matter what career path I end up in.

I would also like to thank my thesis committee, Dr. Dan Kalderon and Dr. Julie Canman for the valuable suggestions and input over the years during my committee meetings. I thank Dr.

Gary Struhl and Dr. Barth Grant for reading my dissertation and for serving on my thesis committee. I would also like to thank my collaborators Dr. Paul Langridge and Dr. Gary Struhl for the work they have done which served as a basis for my second project, the many meetings and their insightful comments on the project. I am grateful to the Department of Biological

Sciences, especially Sarah Kim and Ellie Siddens for their administrative help throughout.

My life in graduate school would look a lot different if not for my wonderful colleagues in the lab. They are some of the smartest, most hard working and most selfless people I have met. I thank Dr. Claudia Tenen and Dr. Yuting Deng for their continuous friendship, encouragement and scientific/technical help along the way; Dr. Ryan Underwood and Dr. Claire de la Cova for their patient guidance and mentorship during my rotation in the lab as well as through the years; my bay mate Hana Littleford for her friendship and scientific discussions along the way. I thank Dr. Michelle Attner, Justin Benavidez, Katherine Luo, Catherine

O’Keefe, Henry Kim, Justin Shaffer, and Dr. Julia Wittes for making my experience at graduate school an enjoyable one. I would also like to thank Gleniza Gomez for her uplifting presence and technical support.

I am beyond grateful for the friends I have met in graduate school. I would like to thank

Dr. Julie Oh and Dr. Yaqiong Chen for their continuous support and encouragement. I am extremely thankful for the friendships outside of graduate school, especially to Stephanie, Rachel and Kony. I can always count on them to brighten up a bad day and to bring a different perspective on things.

I would also like to thank my fiancé, Dr. Alexander Hsieh, for his camaraderie, support and love throughout. I could not imagine how this journey would turn out without him. Finally, I am forever grateful and indebted to my parents, Mark Chan and Qing Qing Liu, for their unconditional love and support even though they are miles away. Without their sacrifice, I would not be where I am today.

vii

Chapter 1: Introduction

1.1 Overview of Notch signaling in C. elegans

1.1.1 Overview of Notch signaling in C. elegans embryogenesis

Notch mediated signaling occurs throughout C. elegans embryogenesis, playing major roles in specifying cell fates at different stages during embryogenesis (Figure 1). The first division of the fertilized egg produces two cells, AB and P1. These cells undergo reproducible patterns of division and differentiation that are referred to as cell lineages (Sulston, 1983). The division of P1 is oriented along the anterior/posterior axis; this division is asymmetric and produces daughters (EMS and P2). The division of AB on the other hand is symmetric and produces anterior and posterior daughters ABa and ABp respectively. The first Notch signaling event occurs as early as 4-cell embryonic stage, where P2 expresses a Notch ligand, APX-1, and signals to ABp, which expresses GLP-1/Notch (Figure 1). This signaling event helps restrict pharyngeal cell fates to the ABa lineage by repressing transcription of pha-4 in ABp, a key transcription factor that is sufficient to induce pharyngeal development. Depleting APX-1 or

GLP-1 at this stage by for example using a temperature sensitive allele of apx-1 or glp-1 results in hyper-induction of pharyngeal tissue, evident by cell fate markers or morphology (Mello et al.,

1994; Mango et al., 1994; Mickey et al., 1996).

Closely related to and functionally redundant with GLP-1 and APX-1, another Notch receptor/ligand pair LIN-12/Notch and LAG-2 plays a major role in later stages of C. elegans embryogenesis (Figure 1). One example is the specification of the excretory cell. Excretory cell forms one of the three unicellular tubes that make up the C. elegans excretory system (Sundaram and Buechner, 2016). An important role of the excretory system is osmoregulation; laser ablation of the excretory cell causes characteristic clear and rod-like larval lethal phenotypes that is

consistent with the loss of osmoregulation function (Nelson and Riddle, 1984; Forrester and

Garriga, 1997). Similarly, loss of LAG-2, or the lin-12 glp-1 double mutant, also results in a rod- like larval lethal phenotype due to the loss of excretory cell (Lambie and Kimble, 1991).

1.1.2 Overview of Notch signaling in C. elegans Germline Development

The C. elegans germline depends on continued production of ligands from the somatic gonad. Early in hermaphrodite larval development, two somatic gonadal cells form at the distal ends of the somatic gonad, these distal tip cells (DTCs) function as migratory cells to extend the gonad arms distally, and also as germline niche cells to maintain proliferation of the germline

(Kimble and Hirsh 1979). The niche function of the distal tip cell is mediated by Notch signaling; DTCs express the DSL ligands LAG-2 and APX-1, activating GLP-1 in the germ cells that are in contact with DTC (Hansen and Schedl, 2013; Kershner et al., 2013). Activation of

GLP-1/Notch maintains the mitotic cell fate of the germline by repression of GLD-1 and GLD-2- mediated meiotic pathway (Crittenden et al, 2002; Hansen et al., 2004; Suh et al., 2006; Brenner and Schedl, 2016). Abolishing Notch ligand in the DTCs or Notch in the germ cells would therefore result in sterility, while constitutive or ectopic activation of GLP-1/Notch in germ cells results in an expansion of mitotic cell population, called a germline tumor (Figure 2) (Seydoux et al., 1990; Pepper et al. 2003).

It was thought that the germline needs to be in direct contact with the DTCs to activate

Notch pathway. Studies have shown using the DTC caps the distal-most ∼3–4 cell diameters and has extensive contact through intercalating cellular processes in a region known as the DTC plexus that extends 8–9 cell diameters from the distal tip, which roughly corresponds with the

mitotic GSCs (Byrd et al., 2014) (Fitzgerald et al. 1994). However, there is no direct evidence for the requirement of these processes in activating GLP-1 in the germ cells. Intriguingly, the male gonad differs from the hermaphrodite gonad drastically in niche architecture and yet the extent and number of mitotic germ cells are comparable between the two sexes. The male DTCs do not have the extensive plexus, and more importantly, there is evidence of transcriptional activation of Notch downstream targets in cells that are not in contact with the male DTCs

(Crittenden and Kimble, 2019). One possible explanation for this is diffusible Notch ligands.

One such ligand, DSL-1, which is known to affect Notch-mediated patterning of the vulva, also affects the number of mitotic germ cells in the male germline (Crittenden and Kimble, 2019).

1.1.3 Notch signaling in the anchor cell (AC)/ventral uterine precursor cell (VU) decision

Within the C. elegans hermaphrodite somatic gonad, Notch plays an important role in specifying a somatic regulatory cell known as the anchor cell, which induces the formation of a vulva, organizes uterine patterning, and orchestrates the uterine-vulval connection. In late L1, four cells in the somatic gonad, two alpha cells and two beta cells, have the potential to become the AC (Kimble and Hirsh, 1979; Kimble, 1981; Seydoux and Greenwald, 1989). Laser ablation experiments suggested that the beta cells lose the potential to be an AC relatively quickly

(Seydoux et al., 1990), and at 20o, do not require lin-12 activity to become VUs (Sallee et al.,

2015). In L2, the alpha cells undergo a LIN-12/Notch-mediated decision so that one cell become the AC and the other becomes a ventral uterine (VU) cell. The cell that has high LIN-12 activity becomes the VU and the cell that express LAG-2 and has low LIN-12 activity becomes the AC

(Figure 3). Removing LIN-12 or LAG-2 genetically results in a 2AC phenotype, which is

detectable either by cell fate markers or cell morphology (Greenwald et al. 1983; Seydoux and

Greenwald, 1989).

Initially, both cells express LIN-12 and LAG-2, and it was originally thought that relative birth-order biases the first-born alpha cell to accumulate more LIN-12 and thus biases that cell towards a VU fate (Wilkinson et al., 1994). However, new study from our lab suggests that lin-

12 expression is initially controlled by hlh-2 and that a the principal event is the relative time of onset of HLH-2 expression in the parents of the alpha cells. This subsequently primes the descendants of the cell with earlier onset of HLH-2 expression to accumulate more LIN-12

(Attner et al., 2019). In combination with a positive feedback loop on lin-12 expression, eventually that cell has a higher LIN-12 activity and becomes the VU (Wilkinson et al, 1994).

1.1.4 Notch signaling in Vulval Precursor Cell specification

C. elegans Vulval Precursor Cells (VPCs) give rise to the 22 cells that make up the vulva.

In early L2, 6 VPCs P3.p to P8.p are equally competent to adopt vulval cell fates. P6.p is specified by EGFR signaling and adopts 1o fate in L3, however ERK activity was observed in early L2, suggesting that the EGFR pathway is activated prior to specification (Hill and

Sternberg, 1992; de la Cova et al., 2017). P6.p then expresses two transmembrane Notch ligands,

LAG-2 and APX-1, and predicted secreted ligand, DSL-1, and signals to neighboring cells P5.p and P7.p, activating the LIN-12/Notch pathway in those cells and adopting 2o cell fate (Chen and

Greenwald, 2004). Cells adopting 1o and 2o cell fate will continue to divide and eventually form a functional vulva in adult, while other VPCs that did not undergo EGFR and Notch signaling will adopt a 3o cell fate, divide once and fuse with the hypodermis (Figure 4) (Sternberg, 2005).

Besides transmembrane Notch ligands, secreted Notch ligands also act on the VPCs to facilitate their specification. The role of secreted ligands will be detailed in section 1.4.

1.2 Overview of Notch signal transduction

1.2.1 Notch signaling pathway

In canonical Notch signaling (Figure 5), both the Notch receptor and its DSL

(Delta/Serrate/Lag-2) ligands are transmembrane proteins with extracellular domains containing

EGF repeats (Gordon et al., 2008). The Notch receptor is cleaved at site 1 by furin during maturation in mammalian cells (Logeat et al., 1998), but not in Drosophila (Kidd and Lieber,

1998). There is currently no direct evidence for S1 cleavage in C. elegans.

Ligand binding initiates proteolytic cleavage of the Notch receptor at site 2 (S2) in the ectodomain and site 3 (S3) in the transmembrane domain (Figure 5). ADAM (a disintegrin and metalloprotease) family metalloproteases, including SUP-17/Kuzbanian and ADM-4/TACE, are responsible for the first cleavage event at site 2 (S2) (Rooke et al., 1996; Pan and Rubin, 1997;

Wen et al., 1997; Brou et al., 2000; Jarriault and Greenwald, 2005). The S2 site is normally concealed within the Notch Negative Regulatory Region (NRR), which includes the LIN-

12/Notch Repeats (LNRs) (Sanchez-Irizarry et al.2004; Gordon et al. 2009). Binding of the DSL ligands to Notch relieves the inhibition of S2 cleavage site and allows for the proteolytic cleavage by the ADAM proteases. Gamma-secretase, an enzyme complex that consists of a

Presenilin (SEL-12 or HOP-1 in C. elegans), APH-2/nicastrin, PEN-2 and APH-1, mediates the second cleavage at site 3 (S3) in the transmembrane domain (Greenwald and Kovall, 2013). S3 cleavage results in translocation of the Notch intracellular domain (NICD) to the nucleus. The

NICD then interacts with LAG-1/CSL (CBF1/Suppressor of Hairless/LAG-1) and SEL-

8/Mastermind-like (MAML) to form a complex and activate the transcription of target genes

(Christensen et al, 1996; Doyle et al., 2000; Petcherski and Kimble, 2000; Struhl, et al, 1993; Wu et al., 2000) (Figure 5).

1.2.2 The domain structure of Delta/Serrate/Lag-2 (DSL) protein ligands for Notch

There are three transmembrane C. elegans DSL proteins, LAG-2, APX-2 and ARG-1, two D. melanogaster DSL proteins, Jagged and Serrate, and six H. sapiens DSL proteins,

Jagged1-2 and Delta-like 1-4. The N-terminal region of the transmembrane ligands, with the exception of C. elegans ligands, contains a conserved ∼100 amino acid MNNL (Module at the

N-terminus of Notch Ligands) domain. Structural studies suggest that this region makes direct contact with Notch EGF-repeats, and mutations in this region can abolish, enhance or diminish

Notch binding in an in vitro setting. Missense mutations associated with Alagille syndrome are mapped to this region in Jagged1, however, MNNL’s role in vivo in other DSL ligands has not yet been supported.

All ligands contain a distinct cysteine-rich module called a DSL domain near the N- terminus, followed by EGF-like repeats that precede the transmembrane domain. These regions are relatively well characterized. In vitro study shows that the DSL domain is necessary but not sufficient for interactions with Notch (Shimizu et al., 1999). Mutations within the DSL domain and deletion of the domain are associated with losses in Notch signaling in both vertebrate and invertebrates (Henderson et al., 1997; Henderson et al., 1994; Morrissette et al., 2001; Parks et al., 2006; Tax et al., 1994; Warthen et al., 2006). In addition, a conserved motif called DOS

(Delta and OSM-11-like proteins) has been proposed to exist within the first two EGF-like

repeats that is proposed to cooperate with the DSL domain (Komatsu et al., 2008). Mutational and structural studies indicate a contributory role for the DOS domain in Notch binding and signaling distinguishing them from the remaining EGF-like repeats (Cordle et al., 2008; Komatsu et al., 2008; Parks et al., 2006; Shimizu et al., 1999). Surprisingly, mammalian Dll4 and Dll3 and all C. elegans DSL ligands lack a DOS motif and it has been proposed that optimal activation of

Notch signaling by DSL domain-only containing ligands requires cooperative Notch binding by

DOS domain containing non-canonical ligands (Komatsu et al., 2008)(Figure 6). The role of these putative non-canonical ligands will be detailed in section 1.4.

While the extracellular domains of DSL are relatively well conserved in their overall domain architecture, there is less conservation within the intracellular domain across species.

With the exception of Dll3 which does not have any intracellular lysines and was thought to be working solely as an inhibitory ligand, all transmembrane DSL ligands contain multiple lysine residues that are potential sites for modification by distinct E3 ubiquitin ligases (outlined in section 4). This ubiquitination is critical for ligands to activate Notch signaling in Drosophila and vertebrates, and the role that this modification plays will be detailed in section 1.3.

1.2.3 Notch extracellular domain architecture and regulators

Both the structure of the Notch receptor and its signaling pathway are well conserved among the vertebrate and non-vertebrate orthologs. There are two C. elegans Notch proteins,

LIN-12 and GLP-1, one D. melanogaster Notch, and four Notch proteins in H. sapiens, Notch

(1-4) (Figure 7). All the orthologs of Notch contains extracellular domains that contain a variable number of EGF repeats, ranging from 10 and 13 in C. elegans GLP-1 and LIN-12 respectively, to 36 in H. sapiens NOTCH1. Theses EGF repeats mediate binding with DSL proteins, and

numerous studies in vertebrates and Drosophila have identified modifications to these domains such as glycosylation that modulate ligand receptor binding (Goode et al., 1996; Moloney et al.,

2000). However, there is limited evidence in C. elegans that such requirement is conserved. The

EGF repeats are followed by three LNR (Lin-12/Notch Repeats) motifs and the heterodimerization domain (in vertebrate proteins). The region from the LNR motifs to the membrane form the NRR (Negative Regulatory Region) and is well characterized by genetic and structural studies. Many mutations that were discovered through genetic screens in C. elegans mapped to this region, and a majority of them have been found to be cause hyperactivity of

Notch in a ligand-independent manner (Greenwald and Seydoux, 1990; Berry et al., 1996?).

Similar and equivalent mutations that are associated with T-cell Acute Lymphoblastic lymphoma

(T-ALL) and other disease are also mapped to this region in mice and cell culture models (Weng et al., 2004).

Structural study of the NRR region of human Notch2 showed that the S2 cleavage site is buried in the NRR because of an extensive interaction surface between the LNR repeats and the heterodimerization domain (HD), which constitutes two subunits tightly entwined in an α-β- sandwich. As a consequence of the cap-like covering of LNR over the HD region, the protein is locked in an autoinhibited mode (Gordon et al, 2007). The next few sections will detail the current model in the field of how such autoinhibition is relieved upon ligand binding.

1.3 Notch activation by DSL

1.3.1 Notch activation by DSL through Epsin-mediated endocytosis

Early studies in Drosophila suggested that trans-endocytosis of Notch ECD by cis- endocytosis of DSL is required for Notch signaling. This process depends on Dynamin, a

GTPase responsible for the scission of newly formed vesicles on the membrane (Seugnet et al.,

1997; Parks et. al., 2000). These studies found a role for endocytosis in Notch activation in both

DSL-expressing cells and Notch-expressing cells.

Several independent studies in both Drosophila and vertebrates later discovered that DSL proteins undergo a specific endocytic pathway to activate Notch. First, they found that E3- ubiquitin ligases Neuralized and Mind bomb interact physically with DSL intracellular domain and promote DSL ubiquitination, and that they act in the DSL-expressing cells for Notch activation in neighboring cells (Itoh et al., 2003; Wang and Struhl, 2005). Mutations of DSL intracellular domain lysines, which are required for ubiquitination, or knocking out Mind bomb or Neuralized renders DSL proteins unable to activate Notch, and produces hallmark Notch phenotypes (Itoh et al., 2003; Wang & Struhl, 2005). Second, other studies found that a different protein, Epsin, is required for Notch activation, and is only required in DSL-expressing cells.

Epsin is an endocytic protein that bind phosphatidylinositol (4,5)-bisphosphate [PtdIns(4,5)P2] in the plasma membrane as well as Clathrin and other accessory protein in coated pits. More importantly, it also contains Ubiquitin Interacting Motifs, leading to the idea that the E3- ubiquitination ligases act upstream of Epsin to promote DSL function (Wang and Struhl, 2004;

Tian et al., 2004; Chen et al., 2009; Xie et al., 2012) (Figure 8).

There were two popular models of why DSL has to undergo endocytosis to signal to

Notch. One model is the “recycling” model, which suggests Epsin is required before ligand- receptor binding. Ligand on the cell surface needs to be endocytosed to be modified and activated, before being recycled back to the cell surface where it can bind receptor. Alternatively,

DSL can be “recycled” to be repositioned to better signal Notch. One study showed that the recycling proteins Sec15 and Rab11 may work in signal-sending cells to promote Delta recycling and thereby signaling (Emery et al., 2005; Jafar-Nejad et al., 2005). In addition, two studies proposed that Neuralized-dependent Delta trafficking from the basolateral membrane to an apical actin-rich structure (transcytosis) juxtaposes Delta with Notch on adjacent cells and thus enables it to signal (Benhra et al., 2010; Rajan et al., 2009).

Another model is the “pulling” model, which suggests that endocytosis simply supplies a pulling force strong enough to dissociate Notch ECD from its ICD. I will further discuss evidence supporting this model in the following sections.

1.3.2 Requirement of force in Notch activation

There are several in vitro studies suggesting the requirement of force in Notch activation

(Stephensen and Avis, 2012; Gordon et al., 2015; Luca et al., 2017; Meloty-Kapella et al., 2012;

Seo et al., 2016). One study developed a high-throughput magnetic tweezers assay to apply a wide range of pN-scale forces to Notch receptors on the cell-surface. They used magnetic tweezers and applies force to cell-surface receptor molecules bound to ligands on paramagnetic beads. By controlling the distance between the cells and the magnet, it is possible to vary the force applied to cells as a function of their well position on the plate. This in vitro magnetic tweezers assay revealed that the isolated activation switch undergoes a transition from protease

resistance to sensitivity between 3.5 and 5.4 pN of force (Gordon et al., 2015). However, these studies do not provide insight on the force requirement on physiological levels, and more importantly, did not preclude the possibility of recycling model since they did not test whether

Epsin mediated endocytosis provides the force necessary for Notch activation.

1.3.3 An In vivo demonstration of pulling force model

A recent paper provided in vivo evidence for the “pulling” model and against the

“recycling” model (Langridge and Struhl, 2017). Specifically, they developed a chimeric Notch ligand and receptor system, by replacing the ectodomain of Notch ligand and receptor with

Follicle Stimulating Hormone (FSH) and part of the ectodomain of its receptor (FSHR). Since

FSH is a secreted molecule and there is also no evidence that FSH needs to be modified to bind its receptor, if modification model is correct, then these FSH constructs would not be expected to overcome the requirement for Epsin. However, such chimeric ligand-receptor pairs were able to recapitulate endogenous Epsin-dependent Notch signaling, therefore arguing against the modification model. The “pulling” model postulates that the NRR opens up when receptor bound ligand is endocytosed. Using the chimeric ligand receptor system, Langridge and Struhl replaced the NRR region of the chimeric receptor with the A2 domain from von Willebrand Factor, a well characterized force sensor, with a cleavage threshold at ~ 8 pN (Zhang et al., 2009). Such chimeric receptor does not result in Notch cleavage. However, when NRR is replaced by a mutant form of A2, with a lower force threshold at ~6 pN (Xu and Springer, 2013), the chimeric receptor is able to recapitulate endogenous Notch activity in an Epsin-dependent manner, suggesting a requirement for Epsin-dependent pulling force in ligand function (Figure 9). This also shows that NRR functions as a force sensor, tuned to a similar range as the mutant A2

domain and that Notch activity can be recapitulated using chimeric receptors with heterologous domains that are cleaved in response to force.

1.4 Role of secreted DSL

1.4.1 Secreted DSL in Drosophila and vertebrates

Early studies in Drosophila tested the function of an artificial DSL lacking the intracellular domain or lacking both the intracellular and transmembrane domain, and found that these ligands lose the ability to trans-activate Notch, but have strong inhibitory interaction with the receptor (Hukriede and Fleming, 1997; Sun and Artavanis-Tsakonas, 1996; Sun and

Artavanis-Tsakonas, 1997). This was later found to be due to a phenomenon called cis- inhibition, likely by sequestration of receptor to preclude its availability to interact with ligands expressed on neighboring cells. This plays in important role in a subset of Notch-dependent developmental contexts (de Celis and Bray, 1997; Jacobsen et al., 1998; Klein and Arias,

1998; Klein et al., 1997), however whether cis-inhibition occurs in C. elegans remain unclear.

Most of the studies are done by artificially manipulating DSL. However, there is no evidence for naturally occurring secreted DSL ligands in both Drosophila and vertebrates.

1.4.2 Secreted DSL in C. elegans

On the other hand, the same manipulation of DSL done in Drosophila has the opposite effect when done in C. elegans. Instead of inhibition, LAG-2 and APX-1 lacking an intracellular domain or lacking both intracellular and transmembrane domain can activate Notch, suggested by their ability to rescue a lag-2 hypomorph and to cause ectopic activation in a Notch receptor- dependent manner (Fitzgerald and Greenwald, 1995, Henderson et al, 1997).

C. elegans has naturally occurring predicted secreted ligands that contain a DSL domain and EGF repeats. Based on sequence, there are 5 predicted secreted DSL ligands, DSL-1, 3,4,5 and 7. In particular, DSL-1 was functionally tested and found to activate Notch in a manner similar to the secreted form of LAG-2 and APX-1, as it can rescue a lag-2 hypomorph as well as cause ectopic activation of Notch (Chen and Greenwald, 2004). While DSL-1 may be able to signal to Notch by itself, in a physiological context it acts in cells where there are other, transmembrane ligands, such as LAG-2 and APX-1 in the VPCs and ARG-1 in the male germline (Chen and Greenwald, 2004, Sallee et al., 2017, Crittenden et al,. 2019).

Proteins that contain only the Delta and OSM-11 (DOS) motif have been proposed to be enhance Notch activity in C. elegans (Figure 8). OSM-11, DOS-1 and OSM-7 in particular are found to be involved in VPC specification, acting synergistically with each other and with other transmembrane DSL ligands and secreted DSL ligands, such as LAG-2 and DSL-1 respectively

(Komatsu et al., 2008).

1.5 Known roles of epn-1/Epsin in C. elegans

There are many studies on the role of Epsin in Notch signaling in other organisms as mentioned in section 1.3. However, in C. elegans Epsin is mostly implicated in contexts not directly related to Notch signaling, with the one exception detailed below.

1.5.1 Previous evidence for requirement of epn-1 in GLP-1 signaling

Given Epsin’s important and distinct role in DSL function in Drosophila and vertebrates, it raises the question whether Epsin also plays the same role in C. elegans Notch signaling. There is some evidence suggesting the only C. elegans Epsin ortholog, epn-1, plays a role in Notch

signaling. Tian et al. (2004) used RNAi to test the effect of epn-1 on GLP-1 signaling in the germline. epn-1 RNAi seemed to reduce germline proliferation only when combined with a glp-

1(bn18), a weak hypomorph. Next, to distinguish whether RNAi acts in the signaling cell, the somatic DTC, or in the signal receiving cell, the germ cells, they performed the same RNAi experiment in a genetic background that renders somatic cells deficient for RNAi, while the germ cells remain sensitive to RNAi. They found that epn-1 RNAi had no effect on the germline in such genetic background, and concluded that epn-1 acts in the somatic cells to mediate Notch signaling, in agreement with the role of Epsin in Drosophila and vertebrate (Tian et al., 2004).

1.5.2 Role of epn-1 in other contexts: embryogenesis and LDLR uptake

Animals defective in epn-1 are lethal and arrest at the embryonic stage, however, the cause of lethality has not been examined closely. One study established a role for epn-1 in programmed cell death in the C. elegans embryo, in particular in the engulfment of apoptotic cells by phagocytes after cell death. Through a genetic screen, Shen et. al. (2013) identified several genes, including epn-1, among other genes in endocytic pathway such as chc-1 and dyn-

1, that are involved in the process. epn-1, chc-1 and dyn-1 acts downstream of the CED pathway to enrich for pseudopod formation by promoting actin assembly. They also generated a null allele of epn-1 and showed that they can rescue its lethality and cell corpse engulfment defect by expressing epn-1 cDNA in cell types that can function as engulfing cells, suggesting the defect in cell corpse engulfment is perhaps the cause or one of the many causes for epn-1(0) lethality

(Shen et al., 2013).

Another context in which epn-1 was shown to be required in C. elegans is low-density lipoprotein receptor (LDLR) internalization. Kang et al. (2013) performed a genome wide RNAi screen to look for gene required in LDLR endocytosis, and identified epn-1 to be one of the factors. Intriguingly, they found that epn-1 lacking the UIM motif can rescue an epn-1 hypomorphic allele as well as a partial rescue of epn-1 LRP-1 trafficking defects (Kang et al.,

2013). However, the cause for epn-1(0) lethality remains unclear, raising the possibility that perhaps defects in Notch signaling may be one of the causes for epn-1 lethality. I investigate this possibility in Chapter 2.

Chapter 1 Figures

Figure 1. Notch signaling in C. elegans embryogenesis. Blue arrows indicate Notch signaling events. The first two Notch signaling events are mediated by maternal apx-1 and glp-1. The first Notch signaling occurs at 4 cell stage in ABp to repress induction of pharyngeal cell fate. The second Notch signaling occurs at 12 cell stage to now induce pharyngeal cell fate. The third and fourth Notch signaling events are mediated zygotically. The third interaction is mediated by LIN- 12 , GLP-1, LAG-2 and APX-1. Activation of Notch here specifies ABplaaa, which is the precursor for all the left head cells. The fourth interaction mediated by LIN-12, GLP-1 and LAG- 2 specifies ABplpapp, which is the precursor for the excretory cell as well as a few rectal cells.

Figure 2. Notch signaling in C. elegans germline development (adapted from Hubbard and Greenstein, 2005). Four cells Z1 to Z4 form the gonad primordium in the embryo. In the L1 stage, Z1 and Z4 will divide and form the somatic gonad, which include regulatory cells such as the DTCs (in red), sheath/spermathecal precursor cells (in light blue) and other regulatory cells such as the AC and VUs (in white). Z2 and Z3 will divide and from the germline. As the animal develops, DTC expresses Notch ligands APX-1 and LAG-2 to activate GLP-1 signaling in the distal germline to promote mitosis (in yellow). Germ cells further away from the DTCs will exit mitosis and enter meiosis (in green) and eventually become oocytes.

Figure 3. Notch signaling in C. elegans AC/VU decision. In L1, Z1 and Z4 begin their lineage to give rise to 12 cells that in the L2 stage that form the developing somatic gonad primordium. The α cells undergo the AC/VU decision during the L2, while the germ cells (in white) proliferate and the distal tip cells (DTCs, in red) guide distal gonad elongation. The α cells (in green) initially express both LIN-12 and LAG-2. A bias in LIN-12 accumulation as well as a LIN-12 positive feedback loop result in one cell with more LIN-12 activity becoming the VU and the other becoming the AC.

Figure 4. Notch signaling in VPC specification. During the L1 stage, the six VPCs, numbered P3.p-P8.p, are born. During the L2 stage, the AC expresses LIN-3/EGF and activates EGFR signaling in P6.p, which is positioned below the AC. In L2 molt, P6.p expresses Notch ligands LAG-2, APX-1 and DSL-1 and adopts 1 o fate in L3. Notch ligands from P6.p activate LIN-12 in neighboring cells P5.p and P7.p, adopting the 2o fate (in blue). The descendants of 1o and 2o VPCs will generate the vulva, and the daughters of the 3o VPCs will fuse to the hypodermis, hyp7.

Figure 5. Overview of Notch signaling. Binding of DSL ligands (in red) to Notch (in blue) relieves the LNR (blue line) auto-inhibition of S2 cleavage site and triggers LIN-12 activation and subsequent S3 cleavage, releasing the LIN-12/Notch intracellular domain (LIN- 12(intra)/NICD). The LIN-12/Notch intracellular domain then translocates to the nucleus, and binds to LAG-1/CSL, SEL-8/MAML, and other coactivators to form a nuclear complex that activates target gene transcription.

Figure 6. DSL in C. elegans, Drosophila and vertebrates (adapted from D’Souza and Weinmaster, 2010). All DSL are conserved in their overall domain organization, all consist of DSL domain and variable number of EGF repeats. The intracellular domain of DSL lack conservation in their sequence, except most ligands contain a predicted PDZ domain. PDZ play a key role in interacting with cytoskeletal components. In known Notch ligands from Drosophila and vertebrates, the DOS motif is always located immediately following the DSL domain. C. elegans DSL-containing Notch ligands on the other hand do not have DOS motifs. In fact, DOS-motif containing proteins exists separately as secreted and transmembrane ligands (not shown here), and are shown to be involved in LIN-12 mediated VPC specification.

Figure 7. Domain organization of C. elegans, Drosophila and human Notch (adapted from Gordon et al., 2008). The extracellular domains of all orthologs of Notch contain a variable number of EGF repeats (in purple), ranging from 13 in C. elegans GLP-1 to 36 in H. sapiens NOTCH1. Theses EGF repeats mediate binding with DSL. Following the EGF repeats, the LNR (lin-12/Notch repeats) and the heterodimerization domain (HD) makes up the negative regulatory region (NRR). The S2 cleavage site is located within the HD, and the S3 cleavage site is located within the transmembrane domain. The Notch intracellular domain mediates binding with coactivators in the nucleus. The RBP-Jkappa-associated module (RAM) and ankyrin repeats mediate interaction with CSL, while proline, glutamic acid, serine, threonine-rich (PEST) domain mediates Notch intracellular domain turnover.

Figure 8. Epsin domain structure and function (adapted from Tessneer et al.) Epsin contains an Epsin N-terminal Homology (ENTH) domain, Ubiquitin Interacting Motif (UIM), aspartate-proline-tryptophan (DPW) repeats and arginine-proline-phenylalanine (NPF) domain. The ENTH domain binds phosphatidylinositol (4,5)-bisphosphate [PtdIns(4,5)P2] in the plasma membrane. The UIM interacts with a ubiquitinated substrate on the membrane. The DPW repeats interact with Clathrin and Clathrin adaptors, while the NPF domain interacts with Eps15 homology (EH) domain proteins, which are often involved in endocytosis.

Figure 9. In vivo force sensor in Drosophila. Langridge and Struhl (2017) developed a chimeric Notch ligand and receptor system, by replacing the ectodomain of Notch ligand and receptor with Follicle Stimulating Hormone (FSH) alpha subunit (in dark red) and ectodomain of its receptor (FSHR) (in dark blue sphere). In presence of FSH beta subunit (in light red), ligand can activate Notch cleavage at S2 site (yellow star) in an Epsin-dependent manner. Another advantage of this ligand receptor system is its modularity. One can replace the NRR region of the chimeric receptor with any sequence that contains a ADAM protease site, for example the A2 domain from von Willabrand Factor, a well characterized force sensor, with a cleavage threshold at ~ 11pN (Zhang et a., 2009). Such chimeric receptor does not result in Notch cleavage. However, when NRR is replaced by a mutant form a A2, with a lower force threshold at ~2pN (Xu and Springer, 2013), the chimeric receptor is able to recapitulate endogenous Notch activity in an Epsin-dependent manner, suggesting a requirement for Epsin-dependent pulling force in ligand function.

Chapter 2: Results

Summary

Epsin-mediated endocytosis is an event required in Drosophila and vertebrates for DSL function in activating Notch S2 proteolytic cleavage. Specifically, when bound to receptor, DSL undergoes Epsin-mediated endocytosis and generates a pulling force that relieves the autoinhibition of Notch S2 cleavage site, and allows for subsequent cleavage events. However, there is limited evidence for the requirement of Epsin-mediated endocytosis in C. elegans Notch signaling. Moreover, C. elegans has naturally occurring secreted ligands, raising the question as to how C. elegans DSL activates Notch and whether Epsin-mediated endocytosis is required.

I took two approaches to investigating the requirement for Epsin in the activity of DSL protein activation of Notch in C. elegans. I first showed that loss of EPN-1 does not cause hallmark phenotypes associated with loss of Notch activity during embryogenesis, indicating that

Epsin is not required for Notch activity in C. elegans. Since Epsin interacts with its cargo proteins via their ubiquitination, I then tested if preventing ubiquitination of the intracellular domain of LAG-2 or APX-1 prevents their function. I found that mutations that eliminate the potential of DSL proteins to be ubiquitinated, or deletion of the entire intracellular domain, does not abrogate the ability of the mutant proteins to function as ligands for activating Notch. I describe additional experiments in C. elegans suggesting that secreted ligands may require association with the external cell membrane and summarize results of Paul Langridge and Gary

Struhl suggesting that the NRR of LIN-12 does not require a strong force to reveal the S2 cleavage site.

In sum, we demonstrated that C. elegans transmembrane DSL LAG-2 and APX-1 function independently of Epsin-mediated endocytosis, and provide evidence that such Epsin-

independence can be attributed to a difference in the Notch Negative Regulatory Regions of C. elegans Notch proteins.

Results

epn-1 is not required in APX-1 mediated signaling in embryogenesis

In C. elegans, epn-1 works in conjunction with various factors such as chc-1, dyn-1, and actin in endocytosis or pseudopod formation (Kang et al., 2013; Shen et al., 2013). epn-1(0) is lethal and has cell corpse engulfment defects as well as LDLR uptake defects (Kang et al., 2013;

Shen et al., 2013). However, the cellular basis of its lethality remains unclear. Given its requirement in Drosophila and vertebrates, it is possible that epn-1 is also required in Notch signaling in C. elegans embryogenesis, and such requirement may cause or contribute to epn-

1(0) lethality.

Both GLP-1 and LIN-12 are required discretely and redundantly at different stages of embryogenesis to endow bilateral asymmetry to the worm (Priess, 2005). At 4 cell stage, apx-1 expression is restricted to P2 via post-transcriptional regulation (Mickey et al., 1996, Tabara et. al., 1999). Maternally expressed APX-1 in P2 activates GLP-1 in ABp to repress pharyngeal cell fate (Figure 1a). Here, we asked if loss of epn-1 affects Notch-mediated specification of ABp. To test epn-1 requirement in APX-1 mediated interaction at 4 cell stage, we performed epn-1 RNAi since maternal transcripts are effectively knocked down by RNAi (Figure 1b). To ensure that epn-1 is expressed at 4 cell stage and that epn-1 can be effectively knocked down at 4 cell stage, we generated an C-terminal endogenous ZF1::GFP tag of epn-1 using CRISPR/Cas9

(Dickinson?). EPN-1::ZF1::GFP is expressed in all four blastomeres at the four cell stage of

embryogenesis, and also in postembryonic cells where Notch signaling occurs, such as in the somatic gonad and VPCs (supplemental figure 1).

As a read-out for ABp cell fate, we assayed the expression of tbx-38p::mCherry-Histone

(Hirsch et al., 2018). Activation of GLP-1 in ABp at the four cell stage leads to repression of tbx-38 expression in ABp descendants, and loss of glp-1 activity at this stage leads to excess tbx-

38-expressing cells derived from ABp. (Good et al., 2004; Hirsch et. al., 2018). As a positive control, we performed apx-1 RNAi, and observed 8/11 embryos had ectopic mCherry expression reflecting the lack of GLP-1 activation at the four cell stage, whereas lacZ(RNAi) negative control embryos had wild-type mCherry expression.

We then performed RNAi on epn-1::ZF1::GFP L4 hermaphrodites. To enrich for individuals that had effective RNAi depletion of EPN-1 at the four cell stage, we examined the progeny of treated L4 hermaphrodites to identify individual embryos at the four cell stage that had lost GFP expression. We incubate the slides with embryos at 25oC for 2 hours, then scored the same embryos later for tbx-38p::mCherry-histone expression. All of these treated embryos had wild-type mCherry expression, suggesting that loss of Epsin did not abrogate APX-1-to-

GLP-1 signaling at the four cell stage of embryogenesis. epn-1 is not required in LAG-2 mediated signaling in excretory cell specification

Unlike the first Notch interaction at 4 cells stage, the third and fourth Notch interaction requires zygotically expressed LAG-2 in signal sending cells (Hutter and Schnabel, 1995; Priess,

2005). In particular, the fourth Notch interaction occurs when LAG-2 expressed on MSap or its descendants signals to LIN-12 and GLP-1 to specify ABplpapp fate, which will eventually give rise to the excretory cell (Hutter and Schnabel, 1995; Moskowitz and Rothman, 1996; Neves and

Priess, 2005). Ablation of signaling cell results in the loss of excretory cell due to the lack of

ABplpapp specification (Figure 2a). Similarly, lag-2(0) single mutants, and lin-12(0) glp-1(0) double mutants, lack an excretory cell (Lambie and Kimble, 1991). The excretory cell marker arIs164[glt-3p::Venus] (Shaye and Greenwald, 2015) reflects excretory cell fate, as it is not expressed in lag-2(0) animals and is expressed in rescued lag-2(0); Ex[lag-2(+)] animals (Figure

2).

epn-1(0) mutants arrest in late embryogenesis with numerous gross anatomical defects

(Shen et al., 2013; Kang et al., 2013). Loss of epn-1 can be effectively rescued by an extrachromosomal array, and maternal and zygotic epn-1(0) individuals identified among progeny of rescued epn-1(0); Ex[epn-1(+)] hermaphrodites through loss of the array during gametogenesis (Shen et al., 2013). We assessed whether the excretory cell is present using arIs164[glt-3p::Venus] and found that 100% of epn-1(0) mutants express the marker (Figure

2b). These observations suggest that epn-1 is likely not required for LAG-2 activation of LIN-12 and/or GLP-1 in ABplpapp.

LAG-2 and APX-1 function does not require intracellular ubiquitination or intracellular domain

Next, we asked if intracellular ubiquitination is required for LAG-2 and APX-1 function, given that it is required for Notch ligand function in Drosophila and vertebrates. We chose to take a cis approach as the most direct way to assess whether ubiquitination is important for ligand function. Before describing our approach, we note that mib-1 and FD107.5 are C. elegans orthologs of Mind bomb and Neuralized, the two RING type E3 ligases that ubiquitinate

Drosophila and vertebrate DSL intracellular domains. In C. elegans, the role of F107.5 is not clear, but mib-1 was reported to be involved in the degradation of Survival of Motor Neuron

(SMN) protein and spermatogenesis (Kwon et al., 2013; Ratliff et al., 2018; Herrara and Starr,

2018). Null alleles of mib-1 are viable and only display spermatogenesis defects in a temperature dependent manner. While the viability of mib-1 null at restrictive temperature alludes to a lack of requirement for mib-1 in Notch mediated embryogenesis, mib-1 loss of function mutants were reported to partially suppress the 0AC phenotype of lin-12(n302gf) and germline phenotype of glp-1(ar202gf), therefore suggesting a possible role of mib-1 in C. elegans Notch signaling

(Ratcliff et al. 2018). It was possible that mib-1 acts redundantly with other Ubiquitin ligases to facilitate Notch signaling, and the cis approach addresses this problem.

lag-2(0) is lethal and has characteristic Lag phenotype as a result of failure to specify the excretory cell and rectal epidermal cells (Lambie and Kimble, 1991). If lysines in the intracellular domain are required, then LAG-2 without lysines would not be able to rescue lag-

2(0). As positive control, we first generated a single copy insertion transgene of lag-2 driven under its genomic regulatory sequence at a specific landing site on LGI using CRISPR-Cas9

(Dickinson et al., 2015). We found that it was able to fully rescue lag-2(0). We then made several mutant derivatives, all as single-copy insertion transgenes in the LGI site to remove any potential variables from differences in conventional arrays or random single-copy insertion in the genome.

When we mutated all the intracellular lysines to arginines, including any lysines in the stop transfer sequence, we found that this form was able to fully rescue lag-2(0) such that viable, fertile adults with wild-type egg-laying ability were obtained (Figure 3). Furthermore, when we removed the intracellular domain of LAG-2 entirely, the truncated protein was also able to fully rescue lag-2(0). These observations indicate that ubiquitination of the intracellular domain is not required for LAG-2 function as a ligand.

We then examined the anchor cell (AC)/ventral uterine precursor cell (VU) cell fate decision in the developing gonad. This decision utilizes a single ligand, LAG-2, and a single

Notch protein, LIN-12. In normal development, two cells each have the potential to be an AC or a VU, and LIN-12-mediated interactions between them result in only one cell becoming the AC.

Loss of either lag-2 or lin-12 results in both cells becoming ACs (Kimble and Hirsh, 1979;

Kimble, 1981; Greenwald et al., 1983; Seydoux and Greenwald, 1989). Rescued hermaphrodites carrying the transgene expressing LAG-2 without the intracellular domain (ΔIC) in a lag-2(0) background have only one AC, indicating that LAG-2 does not require its intracellular domain, and therefore does not require ubiquitination, to mediate the AC/VU decision.

Finally, since previous studies suggested that LAG-2 and APX-1 are functionally redundant (Fitzgerald and Greenwald, 1994; Gao and Kimble, 1995), we also performed the equivalent experiment for APX-1. We generated single copy transgene of apx-1 cDNA driven by the same lag-2 promoter and 3’ UTR used for the lag-2 transgenes to test if the lack of a requirement of intracellular domain also applies to APX-1. While these constructs do not contain lag-2 introns, which could affect apx-1 cDNA expression level, we still observed full rescue from APX-1(KtoR) and APX-1(ΔIC). Similarly, wildtype apx-1 as well as apx-1(ΔIC) can also rescue the 2AC defect of lag-2(0). These results suggest that both LAG-2 and APX-1 function, at least in activating LIN-12 and GLP-1 in the contexts we are looking at, does not require intracellular ubiquitination or their intracellular domains.

Assessment of secreted ligand function: evidence for a role for weak membrane association potentiating ligand activity

C. elegans has naturally occurring secreted ligands such as DSL-1. There is evidence from previous studies that both naturally occurring secreted ligand and artificial secreted ligands can activate LIN-12 and/or GLP-1 (Fitzgerald and Greenwald, 1994; Henderson et al., 1997;

Chen and Greenwald, 2004). In these studies, high copy number "simple arrays" expressing just the ectodomains of LAG-2 or APX-1 can rescue lethality of a strong hypomorph lag-2(s1486), which has a large deletion after the transmembrane domain. These secreted ligands can also cause a tumorous germline phenotype, which is indicative of ectopic GLP-1 activation

(Fitzgerald and Greenwald, 1994; Chen and Greenwald, 2004).

Here we tested whether these secreted ligands can rescue lag-2(0) lethality. The lag-2(0) allele used in this study contains a nonsense mutation before the DSL domain, and is therefore predicted to be a molecular null allele that precludes the possibility of secreted ligands associating with the non-functional LAG-2 protein as in the earlier experiments. We generated single copy transgenes of lag-2 (secreted) and apx-1(secreted) in the same manner described above, but did not observe lag-2(0) rescue. It is possible that without a membrane anchor, secreted ligands diffuse away and the local concentration of ligands is just not high enough to activate Notch. To test this, we generated multicopy version of these transgenes by injecting them as simple arrays, which have multiple copies and likely result in overexpression.

Interestingly, we observed a difference in the ability of secreted LAG-2 and secreted APX-1 to rescue lag-2(0). With secreted APX-1, we found rescue of lag-2(0) lethality and rescue of the

2AC defect (Figure 4a), but with secreted LAG-2, we did not. However, both secreted LAG-2 and secreted APX-1 cause ectopic proximal germline proliferation indicative of ectopic activation GLP-1 in the germline (figure 4b) (Seydoux et al., 1990; McGovern et al., 2009),

similar to what had been described in earlier secreted ligand studies (Fitzgerald and Greenwald,

1994; Chen and Greenwald, 2004).

When we took a closer look at the sequence of secreted APX-1, we found that it may contain a serendipitous GPI linkage site based on computational predictions, which is absent in

LAG-2(secreted) (Eisenhaber et al., 1999; Frankhauser and Maser, 2005; Pierleoni et al., 2008;

Gislason et al., 2019). To test if membrane anchor is required for ligand function, we added the

GPI sequence from efn-1 to lag-2(secreted), and generated single copy transgene by directly injecting the construct into heterozygous lag-2(0) which is balanced by a fluorescent marker. We could see surviving lag-2(0) although not full rescue (Figure 4c). We also observed Muv animals, suggestive of ectopic LIN-12 activation in the VPCs (Figure 4c). epn-1 and ubiquitination independence of C. elegans ligand function can be attributed to a difference in the LIN-12 NRR (Paul Langridge and Gary Struhl)

The key to Notch activation is the relief of NRR mediated autoinhibition of S2 cleavage site. Epsin-mediated endocytosis of DSL generates a pulling force strong enough to open up the

Notch NRR allowing the ADAM protease access to cleave at the S2 site. Given the epn-1 independence of C. elegans DSL, we hypothesized that C. elegans Notch NRR differs in conformation such that less or no pulling force is required for Notch activation.

To test this, we collaborated with Paul Langridge and Gary Struhl, who developed an in vivo pulling force sensor (Refer to section 1.3.3 for details). The force sensor is a transmembrane chimeric protein containing a Notch intracellular domain, a Follicle Stimulating Hormone

Receptor (FSHR) ectodomain. In between the ectodomain and transmembrane domain, the protein sequence of one’s choice that contains a membrane protease cleavage site. With the

Drosophila Notch NRR sequence, the cleavage is ligand dependent, and more importantly,

Epsin-dependent (Langridge and Struhl, 2017). However, with the C. elegans LIN-12 NRR sequence, the cleavage is no longer Epsin-dependent, but remains ligand-dependent. This indicates that a difference in NRR can explain the epn-1 and ubiquitination independence of C. elegans DSL function. In addition, secreted and GPI-anchored form of FSH were tested for their ability to activate the chimeric receptor with LIN-12 NRR. They found that the secreted form of

FSH cannot activate the chimeric receptor, while the GPI-anchored form can, which is consistent with what we found in C. elegans

Chapter 2 Figures

Figure 1. epn-1 is not required for GLP-1 signaling at 4 cell stage. A) tbx-38 expression reports APX-1 activity at 4 cell stage. The first division of the fertilized egg produces two cells called AB and P1. Next round of division of P1 is asymmetric and produces daughters EMS and P2, which express distinct sets of proteins. Only P2 expresses APX- 1, represented in yellow. Symmetric division of AB produces ABa and ABp which are initially equivalent and both express GLP-1, represented in blue. However, ABp is positioned to directly contact ligand expressing P2, thus activating GLP-1 signaling in ABp. This first Notch signaling event prevents the later expression of T-box transcription factors such as tbx-38, represented in light red in ABa. In normal development, only ABa descendants express tbx-38, represented in red in ABp. However, in an apx-1 or glp-1 mutant, both ABa and ABp descendants would express tbx-38. B) epn-1 knockdown at 4 cell stage does not affect tbx-38 expression. To test the requirement for epn-1 for APX-1 function at the 4 cell stage, we used a reporter strain that contains a ubiquitous green histone marker, a tbx-38p::mCherry-histone marker, and an endogenously GFP tagged EPN-1, present at the membrane. Top panels show endogenous EPN-1::ZF1::GFP expression with and without epn-1 RNAi. Nucleus is marked by HIS-72::GFP. Endogenous EPN-1::ZF1::GFP is expressed in all 4 cells at 4 cell stage, present in puncta in the cytoplasm as

well as the plasma membrane. epn-1::zf1::GFP can be knocked down by RNAi at 4 cell stage, as membrane and cytoplasmic GFP is no longer visible after RNAi. Embryos were first scored for EPN-1::ZF1::GFP knockdown. The same embryos were then scored at ~150 cell stage when tbx-38p::mCherry-histone is visible, shown in the bottom panels. While there is ectopic mCherry expression in the positive control (apx-1 RNAi), there is no difference in mCherry expression in epn-1 and lacZ RNAi-treated animals. Bottom graph shows quantification of % mCherry positive nuclei/ GFP nuclei for each embryo treated with apx-1, lacZ and epn-1 RNAi respectively (see Methods).

Figure 2. epn-1 is not required for lag-2 function in specifying excretory cell. A) LAG-2 mediated specification of excretory cell. Excretory cell specification occurs in the embryo, and involves an AB descendant called ABplapp which expresses Notch (indicated in blue). Cell ablation and mutant analysis of Notch pathway showed ABplpapp is specified through zygotic Notch signaling, induced by MSap or its descendants which express lag-2 indicated in red. In normal development, one of the ABplapp descendants give rise to the excretory cell, indicated in green, which is missing in lag-2 mutant in L1.

B) epn-1 loss does not affect ABplpapp specification. epn-1(en47) is a null allele with an early nonsense mutation. Yellow excretory cell marker arIs164[glt-3p::Venus] is still visible in epn-1(0) shown in the left panels, as opposed to in lag-2(0) animals where the marker is clearly absent shown in the second to the right panel. epn-1(0) animals were picked based on the lack of enEx[epn-1::GFP] rescuing array, which is expressed ubiquitously in the animal (second to the left panel). The lack of excretory cell is rescued with a lag-2(+) array, which is marked by ceh-22p::GFP (right panel). Ns are shown in table C.

Figure 3. Ubiquitination is not required for LAG-2 function in embryogenesis. A) Strategy for testing LAG-2 and APX-1 function in rescuing lag-2(0). Schematic of single copy insertions of lag-2 and apx-1 constructs (not to scale). Transgenes are tested for their ability to rescue lag-2(0), which is 100% rod-like lethal. For most transgenes, we generate single copy insertion of different transgenes inserted on the same site on LGI using CRISPR-Cas9 (Dickinson et al., 2015). All transgenes contain regulatory sequences from lag-2: a 3.3kb lag-2 promoter as well as a 1 kb lag-2 ‘UTR. For lag-2 based transgenes, we used the lag-2 genomic DNA which includes introns. For apx-1 based transgenes, we used apx-1 cDNA. B) Constructs for testing LAG-2 and APX-1 function in rescuing lag-2(0). lag-2(KtoR) contains 10 lysine to arginine mutations after the transmembrane domain. lag-2(ΔIC) was generated by replacing a fragment from 311aa to 403aa with a stop codon, which results in a protein truncated at 3aa after the transmembrane domain. apx-1(KtoR) contains 4 lysine to arginine mutations after the transmembrane domain. apx-1(ΔIC) generated by replacing a fragment from 417aa to 516aa with a stop codon, which results in a protein truncated at 6aa after the transmembrane domain. lag-2(secreted) has a stop codon replacing 282aa to 402aa, which results in a secreted protein. apx-1(secreted) has a stop codon replacing 403aa to 516aa, which theoretically results in a secreted protein. C) lag-2(0) lethality and 2AC rescue by lag-2 and apx-1 single copy inserted transgenes. *2 AC rescue data is obtained by scoring an AC marker arIs51[cdh-3p::GFP] with the transgene and lag-2(0) in the background. ** Due to lethality of lag-2(0), the data is based on 2 AC phenotype from lag-2(q420ts) at restrictive temperature post L1 from Chen and Greenwald, 2004. ***Due to lethality of lag-2(0), progeny from lag-2(0)/ tmC16 were scored.

Figure 4. Membrane anchor is required for DSL ligand function A) lag-2(0) lethality, 2AC rescue and ectopic germline GLP-1 activation by lag-2 and apx-1 multicopy transgenes. See Materials and Method for position of truncation to generate lag-2(secreted) and apx-1(secreted). *2 AC rescue data is obtained by scoring an AC marker arIs51[cdh-3p::GFP] with the transgene and lag-2(0) in the background. ** Since lag-2(secreted) arrays did not rescue lag-2(0), strains were maintained in a lag- 2(0)/TmC16 background. Ectopic GLP-1 activation in the germline were scored in the same background. Multicopy transgene of apx-1(secreted) can partially rescue lag-2(0), likely due to the presence of a serendipitous GPI site created by the truncation. B) DSL ligand function requires a membrane anchor. Single copy insertion of lag- 2(secreted) or apx-1(secreted) cannot rescue lag-2(0). Consistent with the hypothesis that lag-2(0) rescue by multicopy apx-1(secreted) is due to presence of GPI linkage, addition of efn-1 GPI site sequence to lag-2(secreted) creates a single copy transgene that can now rescue lag-2(0). C) Ectopic GLP-1 activation in proximal gonad. On the left, a microphotograph of an adult animal with tumorous germline as a result of ectopic GLP-1 activation. The strain’s genotype is lag-2(0) rescued by apx-1(secreted) multicopy transgene. On the right, a microphotograph of a wildtype N2 adult germline, with fertilized eggs and oocytes in proximal gonad. White arrows indicate vulvae.

Supplemental figure 1) epn-1::zf1::GFP expression in L3. GFP is observed in cells involved in Notch signaling such as AC (indicated by arrow) and VPCs (P6.p underlined), accumulating in cytoplasm and in puncta on the membrane.

Supplemental figure 2 A) 2AC rescue of lag-2(random peptide). Despite the ability to rescue lag-2(0) lethality, both lag-2(Random Peptide no K) and lag-2(Random Peptide) can only partially rescue the 2AC lethality, while lag-2 without an intracellular domain can fully rescue the 2AC defect. *Due to lethality of lag-2(0), the data is based on 2 AC phenotype from lag-2(q420ts) at restrictive temperature (Chen and Greenwald, 2004). B) Images of 1AC and 2AC from lag-2(q411); arSi62[lag-2(Random Peptideno K)]. White arrows indicate anchor cells marked by arIs51[cdh-3::GFP].

Chapter 3: Discussion

3. Discussion

Here we showed that C. elegans DSL proteins function in an Epsin-independent manner, and that C. elegans transmembrane DSL proteins LAG-2 and APX-1 can function without ubiquitination of intracellular lysines, or even without the intracellular domain. However, their function does appear to require some sort of association with the membrane, as can be provided artificially by an anchor like GPI. These observations contrast with the mechanism found in

Drosophila and vertebrates, that Epsin-mediated endocytosis is required for DSL function. Our collaboration with Langridge and Struhl showed that this difference can be attributed to the difference in the Notch NRRs. However, many questions remain. One immediate question is whether the Epsin-independence of C. elegans DSL function is due to the lack of a requirement, or a lower threshold for a pulling force. I will address them in this section.

3.1 Requirement of pulling force in C. elegans Notch activation

Although we showed that transmembrane DSL LAG-2 and APX-1 do not require Epsin, ubiquitination or intracellular domain to function, we did not show that pulling force is not required for Notch cleavage. There are two questions remained to be answered here: 1) Is pulling force from signal sending cell required? 2) If so, how is pulling force generated if not by Epsin- mediated endocytosis? We attempted to address the first question by testing the function of secreted ligands. We observed lethality rescue with APX-1(secreted), similar to what was found in previous study (Fitzgerald and Greenwald, 1994). However, here we showed that the rescue is likely due to the presence of a fortuitous GPI, suggesting that some sort of membrane anchor is required for ligand function, at least in the context we looked at. This implies that a level of force similar to that generated by a GPI anchor is required for ligand function. Alternatively, GPI anchorage can increase local concentration or proximity of ligand to Notch.

In the Notch cleavage assay used in this study, FSH ligand receptor pair was used for their monomeric nature. In Langridge and Struhl (2017), several other obligate homodimers and monomers (ligand/receptor binding domains of Neurotrophin-3 (NTF)/Tropomyosin receptor kinase C, and GFP/GFP nanobody) were also used to replace the ectodomain of the receptor- ligand pair, and were shown to recapitulate Notch activity (Langridge and Struhl, 2017). Since such chimeric ligand receptor pairs can recapitulate Notch signaling, it suggests that in

Drosophila monomeric Delta-Notch interaction is sufficient for activity. One could conceivably develop a similar assay in C. elegans to ensure monomeric ligand-receptor interaction.

Alternatively, we can begin to infer the requirement of pulling force in signal sending cell by simply visualizing DSL and Notch localization. In Langridge and Struhl (2017), mCherry was inserted in the ectodomain of FSHR-Notch chimeric receptor. Early endosomes were marked by

YFP-Rab5. Normally, FSH-Dl chimeric ligand trans-endocytose the ectodomain of the chimeric receptor into the signal sending cell. In contrast, all mutated forms of ligand that cannot undergo

Epsin/Clathrin pathway fail to show trans-endocytosis of the receptor ectodomain in signal sending cell. We can perform similar experiments in C. elegans by endogenously inserting a fluorescent tag in the ectodomain of LIN-12 or GLP-1, and see if there is a population of receptor that are trans-endocytosed into signal sending cell, which can be marked by membrane markers. While not conclusive, trans-endocytosis of LIN-12 or GLP-1 ectodomain into signal sending cell would support that idea that ligand generates some level of pulling force during

Notch cleavage. However, this experiment would only be possible if the fluorescent tag insertion does not interfere with endogenous Notch activity.

3.2 Alternative Epsin-independent pulling force mechanisms: Clathrin-mediated endocytosis or Clathrin-independent endocytosis of ligand

To address the second question of how pulling force is generated, one could look to the various mechanisms by which membrane proteins are internalized. The best studied route of uptake is via the Clathrin-mediated endocytosis (CME). Clathrin assembles into a lattice structure that deforms the membrane into a bud. The main structure units of clathrin cage are the clathrin heavy chain (chc-1) and clathrin light chain (clic-1). Clathrin does not bind directly to cargo, but to various clathrin adaptors, such as Epsin, AP2 adaptor complex, and others. Once the clathrin coated bud is formed, dynamin (dyn-1) forms spirals around the neck of the clathrin coated pit and promote the fission of the vesicle from the membrane. Subsequently, clathrin is release from the vesicle, a process mediated by chaperone hsc70 and the DNA-J domain co- chaperone auxilin (dnj-25). RNAi and temperature sensitive allele of chc-1 indicates that it is essential for endocytosis and viability in C. elegans (Sato et al., 2009). A temperature sensitive allele of dyn-1 indicated that maternal or early zygotic dynamin function is essential for viability, and was later found to regulate anterior cell polarity in early embryo as well as cell corpse engulfment in late embryo (Clark et al., 1997; Yu et al., 2006; Nakayama et al., 2009). RNAi of dnj-25 results in defects in yolk protein uptake in oocyte (Greener et al., 2001).

On the other hand, Clathrin-independent endocytosis (CIE) is less studied in C. elegans.

CIE pathways can largely be divided into two categories: the large-scale invaginations such as phagocytosis and pinocytosis, and small scale invaginations. One known small-scale CIE pathway in C. elegans is caveolin-dependent endocytosis. Caveolae are small invaginations located on the membrane, often enriched in lipid-raft associated molecules such as cholesterol, sphingolipids, glycosphingolipids and other signaling and receptor proteins (Parton and Simons,

2007). Various molecules such as GPI-anchored proteins, glycosphingolipids and SV40 virus have been shown to be trafficked through CIE (Marsh and Heleniu, 2006; Mayor and Pagano,

2007). Caveolin proteins are required for caveolae formation. C. elegans has two caveolin genes, cav-1 and cav-2. cav-1 has been implicated in meiotic germline progression and neurotransmission (Scheel et al., 1999; Parker et al., 2007). cav-2 has been implicated in trafficking of yolk proteins in the intestine (Parker et al., 2009). However, it remains unknown whether cav-1 and cav-2 double mutant is viable.

With existing tools, one could test the requirement for these factors in DSL function.

Pleiotropy is a major issue in interpreting results from these experiments. These essential genes are known to be involved in multiple cellular processes and contexts. One could achieve temporal and some spatial specificity for example by looking at the effect of a temperature sensitive allele [for example dyn-1(ky51) and chc-1(b1015ts)] using thermogenetics (Hirsch et al., 2018). Another issue is that these factors could also be important for receptor activity or stability or localization. Therefore, it is important to be able to manipulate these factors specifically in signal sending cells. One could do mosaic analysis with a dyn-1 or chc-1 recusing array in the respective hypomorphic background. One could also perform tissue specific RNAi, perhaps against cav-1 and cav-2 which do not have characterized null or temperature sensitive alleles (for an improved tissue specific RNAi experimental design please refer to Appendix). A negative result could simple be due to insufficient knockdown, and a positive result from these experiments would suggest a role for endocytosis in ligand function. These experiments can be a good start for answering the question of whether these endocytic factors are required in signal sending cell for Notch activation. However, a major caveat with these experiments is that it is hard to first, distinguish between non-ligand specific endocytosis and ligand specific

endocytosis, and second, distinguish between endocytosis that is essential for ligand function

(i.e. generating a pulling force) or simply for protein turnover. It is possible that non-specific ligand endocytosis generates sufficient pulling force for ligand function. If we find that there are other endocytic pathways by which DSL is required to function, an interesting question is whether these routes of internalization generate a different level of force than Epsin-mediated endocytosis.

3.3 Alternative Epsin-independent pulling force mechanisms: Receptor endocytosis

The requirement for endocytosis of Notch ligands is well characterized. However, the requirement for internalization of the receptor in Notch activation is less well established. There is evidence in mammalian cells for Notch endocytosis promoting attenuation of Notch signal by expression of the receptor on cell surface, but not its activation (Sorensen and Conner, 2010). On the other hand, there is some evidence in Drosophila that supports a role for receptor endocytosis in Notch activation. Studies found that removing dynamin in signal receiving cells disrupted

Notch activation (Seugnet et al., 1997). Notch can also be targeted for the degradation pathway via ubiquitination by E3 ligases such as Nedd4, and undergoes endocytosis mediated by Numb

(Sakata et al., 2004). Recently, visualization of ligand and receptor in Drosophila shows that ligand-receptor pairs that did not result in Notch cleavage, for example in the case where ligand cannot undergo Epsin-mediated endocytosis, are often endocytosed into the signal receiving cell

(Langridge and Struhl, 2017). This suggests that Notch cleavage is dependent not only on ligand endocytosis, but is a result of competition between ligand endocytosis and receptor endocytosis.

In C. elegans, there is limited evidence for receptor endocytosis being required for Notch activation. Studies have shown that E3 ligases such as ego-2, alx-1, wwp-1 and Rho GTPases such as cdc-42, which are involved in endocytic pathways, can physically or genetically interact

with Notch (Shaye and Greenwald, 2004; Liu and Maine, 2007; Choi and Greenwald, 2010).

However, receptor internalization per se has not been tested for C. elegans Notch signaling. It is possible that the force generated from receptor internationalize is sufficient for activation, in absence of ligand endocytosis, as long as ligands are anchored in some way (discussed below in

3.4). One could test the requirement for internalization factors in Notch activation in signal receiving cells in a similar manner discussed above (chapter 3.2).

3.4 Alternative Epsin-independent pulling force mechanisms: C. elegans Notch activation by ligand-receptor oligomerization or ligand-membrane association

Another model for Notch activation in C. elegans, not mutually exclusive from the models mentioned above, could be that Notch cleavage doesn’t require a pulling force from signal sending cell per se, but simply by force from ligand-receptor oligomerization. Ligand- receptor oligomerization plays an important role in several other signaling pathways such as

GPCRs, RTKs and integrins. There are in vitro and in vivo studies suggest that mammalian and

Drosophila Notch does not require ligand-receptor oligomerization for Notch cleavage (Vooijs et al., 2004; Gordon et al., 2015; Seo et al., 2016; Langridge and Struhl, 2017). However, there is recent evidence that ligand-receptor clustering can attenuate Notch signaling (Nandagopal et a.,

2018). On the other hand, in C. elegans there is genetic evidence for Notch activation from receptor oligomerization. A hypermorphic allele of lin-12 can trans-activate a lin-12 hypomorphic allele (Seydoux and Greenwald, 1990). In addition, LIN-12 without an extracellular domain is only active in presence of wildtype LIN-12 (Katic and Greenwald, 2005).

Another model for Notch cleavage in C. elegans is that the force of ligand-membrane association, along with the association with receptor, is sufficient for Notch cleavage. This is

supported by results from this study, where we showed DSL tethered by a transmembrane domain or by GPI is sufficient for its function.

3.5 Evolutionary and structural basis for pulling force requirement in Notch activation

The lack of conservation in the requirement for Epsin-mediated endocytosis in C. elegans raises the question of whether the Epsin-independence of C. elegans Notch activation represents an ancestral state of Notch signaling. Components for Notch signaling are present in Eukaryotes, with a strong evolutionary conservation of Notch pathway in bilaterians. Some but not all components of the Notch signaling system, such as the receptor, ligand, ADAM protease and gamma-secretase, can be found in basal phyla such as Cnidaria, Placozoa and Porifera.

Intriguingly, these genomes lack sequences with diagnostic domains for Mindbomb and

Neuralized (Gazave et al., 2009). While this could be due to prediction errors or gaps in genome sequenced, it could also suggest that these genes are not a core component of Notch signaling in these organisms, either due to a lack of requirement for Epsin-mediated endocytosis, or a redundancy for E3 ligases. DSL ligand domain analysis of these organisms also predicts that they contain secreted DSL (Figure 1) (Gazave et al., 2009). As these predictions could be due to mis-annotation or assembly of the genomes, the only way to test this is by functional studies, which is scarce in non-bilaterians. There are a few functional studies of Notch and its ligands in

Nematostella, Hydra, and Amphimedon (Kasbauer et al., 2007; Richards et al., 2008; ;

Srivastavas et al., 2008; Richards and Degnan, 2012; Munder et al., 2013; Marlow et al., 2013;

Ringrose et al., 2013.

We can also turn to the receptor for insight, since receptors and ligands often coevolve together (Moyle et al., 1994). In this study, we showed that a difference in NRR sequence can explain the Epsin-independence of C. elegans Notch activation. Structural studies of human

Notch2 reveals that the S2 cleavage site is buried in a small hydrophobic pocket in homodimerization domain (HD) and a “plug” that fills the pocket and block access to the cleavage site is the side chain of Leu1457 on LNR. Leu1457 is fixed in the pocket by hydrogen bonds with other residues on the HD (Gordon et al., 2015). Interestingly, in C. elegans LIN-12, the highly conserved Leu1457 plug that masks the S2 site in the NRR of NOTCH2 is replaced by an arginine residue that probably forms a protective salt bridge with a glutamate residue near its presumed S2 site (Gordon et al., 2015). This leads to the hypothesis that such difference could contribute to the Epsin-independence of Notch activation in C. elegans. Alignment of Notch protein sequence shows that most Caenorhabditis lacks the conserved Leu, as well as representative species across Nematoda and Amphimedon. However, Hydra and Nematostella seems to have the conserved Leu (Figure 2). With the modular nature of the cleavage reporter used in this study, one could test the various NRRs, or mutate residue on human or Drosophila

NRR to probe the significance of these residues in S2 cleavage. This is currently being tested by

Paul Langridge and Gary Struhl.

3.6 Notch signaling in different development contexts in C. elegans

Here we tested wildtype and mutant LAG-2 and APX-1 function in several Notch mediated developmental processes: embryogenesis, AC specification, and germline proliferation.

There are fundamental differences in the role of Notch signaling in these processes. Notch in embryogenesis and germline proliferation are mediated by inductive signaling, in which one cell type often express DSL and signals to another distinct cell type to promote differentiation or proliferation. On the other hand, AC specification is mediated by lateral signaling in which two cells of the same type express both ligand and receptor, and small differences are amplified by feedback loops so that each cell adopt distinct cell fates.

Results from this study suggests that there is different requirement or regulation on ligand in these different contexts. We generated a lag-2 mutant in which we replaced the intracellular domain with a random peptide. The same random peptide was used in Drosophila to test the nature of ubiquitination of DSL; ligand with random peptide that has no lysines cannot activate

Notch and its function restored as long as there are lysines. Both forms of mutant in C. elegans can rescue lag-2(0) lethality, but both exhibit incomplete rescue of the 2AC phenotype, in contrast to lag-2(+) and lag-2(KtoR) where they can fully rescue both types of defects

(Supplemental Figure 2). We hypothesize that there is some sequence within the intracellular domain that is only required in lateral signaling in AC specification, not inductive signaling in an ubiquitin-independent manner. However, lag-2(ΔIC) can also fully rescue both types of defects, suggesting that there isn’t a sequence within LAG-2 intracellular domain that can explain the random peptide phenotype, rather the phenotype we see is likely due to the nature of the random peptide itself. Despite that, this result still indicates a difference in regulation of LAG-2 in the two developmental contexts, perhaps on a biophysical level which is hard to identify using a genetic approach.

We also noticed a difference in the ability of secreted ligand activating Notch in the contexts we looked at. Neither LAG-2(secreted) or DSL-1 can rescue lag-2(0) lethality, but they can activate GLP-1 in the germline and cause a tumorous germline phenotype. This may be due to a difference in receptor activity threshold that is required to activate downstream pathway.

Chapter 3 Figures

Figure 1. Predicted DSL in Bilateria and non-Bilateria (from Gazave et al., 2009). This figure shows the predicted DSL based on Delta domain organization in Bilateria (in blue letters) and non-Bilateria (in green letters).

Figure 2. Alignment of NRR sequences from different Notch proteins. Leu1457 in HmNotch2 and the equivalent residue in other species is highlighted in red box. The leucine residue is conserved in all four human Notch proteins. Human Notch2, (Hs) Drosophila (Dm), Hydra (Hv), and Nematostella (Nv) have the Leucine residue at the position, while C. elegans (Ce), other nematodes (C. briggsae (Cb), M. incognita (Mi) , P. pacificus (Pp), B. malayi (Bm)), and A. queenslandica (Aq) do not. Multiple Sequence Alignment was done using Clustal Omega on default setting.

Chapter 4: Materials and Methods

STAR Methods

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Iva Greenwald ([email protected])

Experimental Model and Subject Details

C. elegans Strains

Strain names and full genotypes are listed in Key Resources Table. Caenorhabditis elegans strain N2 and lag-2(q411)/tmC16[unc-60(tmIs1237)] in which transgenes are generated by microinjection. lag-2(q411) is a null allele, which is homozygous lethal. For injection purposes, it was maintained as a heterozygote using tmC16. tmC16 was obtained from

Caenorhabditis Genetics Center. All strains are maintained at 25oC.

enEx867[epn-1::GFP] is an extrachromosomal array containing epn-1 c-terminally tagged by GFP, driven under the epn-1 genomic sequence. This transgene can rescue the lethality of a null allele of epn-1, epn-1(en47). The rescued strain was a gift from Zheng Zhou.

zuIs178 ubiquitously marks the nucleus of embryos by GFP tagged histone, stIs10138 marks ABp descendants in the embryo that has activated GLP-1/Notch with histone tagged by mCherry, which is driven under the tbx-38 promoter, and faithfully reports GLP-1 activity in the

4 cell stage. These markers were gifts from Julie Canman.

arIs51[cdh-3p::gfp] is expressed in the AC in L2 and L3 and is used as an AC marker.

arIs164[glt-3p::Venus] marks the excretory cell and is visible from L1 to adulthood.

Additional transgenes generated during the course of this study are described in the

Method Details below.

Method Details

Plasmid Generation and Transgenesis for lag-2(0) rescue

All transgenes used to assay lag-2(0) rescue are generated by single-copy insertion on the ttTi4348 locus on LGI using CRISPR-Cas9, or by multicopy simple array. All plasmids used to generate both kinds of transgenes were made by cloning the desired insert into pWZ111 backbone using Gibson Assembly, with pWZ111 cut by SpeI and AvrII. All inserts were generated by PCR with AccurpimeFx. Cut pWZ111 contains the homology arms used for insertion of transgene at LGI site ttTi4348, and a Self-Excising Cassette containing sqt-1p::SQT-

1 and rps-0p::hygG flanked by loxN sites, and a hsp::Cre. Unless heat-shocked, any transgene generated from such pWZ111 backbone would result in rollers and hygromycin resistant animals.

All single copy transgenes were injected into N2 with pAP82 at 50 ng/ul, pGH8 at 10 ng/ul, pCFJ90 at 5 ng/ul, and plasmid at 10 ng/ul.

All extrachromosomal simple arrays were injected into lag-2 (q411) /tmC16 with plasmid at 50 ng/ul and pBS at 100 ng/ul, or plasmid at 25ng/ul and pBS at 125 ng/ul (for details see below).

arSi48[lag-2p::lag-2(+)] is a single-copy insertion on the LGI site generated from pJC70, which contains the lag-2 gDNA sequence. It contains a 3.3kb 5’ upstream sequence of lag-2 as promoter, driving the expression of lag-2 1.5 kb gDNA and 1 kb of lag-2 3’ UTR.

arSi49[lag-2p::lag-2(KtoR)] is a single copy insertion on the LGI site generated from pJC71, with all the intracellular lysines (10 lysines after 307aa) mutated to arginine.

arSi89[lag-2p::lag-2(ΔIC)] is a single copy insertion on the LGI site generated from pJC106, with a stop codon replacing a fragment from 311aa to 403aa, which results in a protein truncated at 3aa after the transmembrane domain. The 3aa after TM are mutated from KYK to

RYR.

arSi88[lag-2p::apx-1(FL)] is a single copy insertion on the LGI site generated from pJC100. It contains the 3.3kb lag-2 promoter, the apx-1 cDNA, and the 1kb 3’ UTR of lag-2.

arSi90[lag-2p::apx-1(KtoR)] is a single copy insertion on the LGI site generated from pJC120 with all the intracellular lysines (4 lysines) mutated to arginine.

arSi91[lag-2p::apx-1(ΔIC)] is a single copy insertion on the LGI site generated from pJC121, with a stop codon replacing a fragment from 417aa to 516aa, which results in a protein truncated at 6aa after the transmembrane domain. The 6aa after TM are mutated from SFSKWK to SFSRWR.

arSi95[lag-2p::lag-2(secreted)-GPI] is a single copy insertion on the LGI site generated from pJC119. lag-2(secreted) is based on the design of the same transgene from Fitzgerald and

Greenwald, 1994, with last 20aa GPI sequence from efn-1 in the C-terminus.

arEx2523, 2530, 2531 [lag-2p::apx-1(secreted)] are simple arrays generated from pJC109 based on the design of the same transgene from Fitzgerald and Greenwald, 1994, with a stop codon replacing 403aa to 516aa, which results in a secreted protein. pJC109 were injected at 50ng/ul.

arEx2524,2525,2526 [lag-2p::lag-2(secreted)] are simple arrays generated from pJC99, which is based on the design of the same transgene from Fitzgerald and Greenwald, 1994. pJC99 was injected at 50 ng/ul for to generate these transgenes.

arEx2527,2528,2529[lag-2p::dsl-1] are simple arrays generated form pJC115, which was injected at 50 ng/ul.

arEx2533, 2534,2535 [lag-2p::lag-2(secreted)::GPI] are simple arrays generated from pJC119. GPI sequence was from the last 20aa from efn-1 sequence. pJC119 was injected at 25 ng/ul.

arEx2536, 2537,2538 [lag-2p::lag-2(secreted)] are simple arrays generated from pJC99, injected at 25 ng/ul.

Assessment of lag-2(0) lethality rescue

To test the single copy transgenes for lag-2(0) lethality rescue, we first generated an intermediate strain with transgenes balanced by oxTi559; lag-2(q411)/tmC16. Transgenes were considered to be rescuing if the strain produced surviving progeny lacking tmC16, meaning they are homozygous for lag-2 (q411). For transgenes that can rescue lag-2(0), we will assay the extent of lag-2(0) rescue, 10-15 egg-laying adult P0s were picked onto plates seeded with OP50.

P0s were removed 3 hours later. Progeny were scored for lethality 24 hours later. All experiments were performed at 25. All strains were PCR genotyped and sequenced to confirm its genotype. All single copy transgenes were assayed and genotyped in the manner described above, with the exception of lag-2(secreted)::GPI, due to lab shut down.

To test the extrachromosomal simple arrays for lag-2(0) lethality rescue, the plasmid was directly injected into lag-2(q411)/TcM16, which were then placed at 25oC. The plasmid backbone contains selection markers such as the sqt-1 gene and hygromycin resistance gene. 4 days after injection, 500 ml of Hygromycin was added to the plates. 8 days after injection, roller animals were picked on to single plates. Transgenic arrays were considered to be rescuing if the roller P0 produced surviving rolling animals lacking TcM16 meaning they are homozygous for lag-2 (q411).

Assessment of lag-2(0) 2AC rescue

To test for all transgenes rescue for lag-2(0) 2AC phenotype, arIs51[cdh-3p::GFP] was crossed into the strains. In normal development, arIs51[cdh-3p::GFP] is expressed in the AC of the somatic gonad primordium in the L2 stage. It remains restricted to the AC until it expands to multiple cells of the utse in L4. In this study, only L3 animals were scored for AC phenotypes and were picked based on VPC and gonad progression. Scoring and imagining of the GFP fluorescence were done at either 40x or 63x with a Zeiss Axio Imager D1 microscope with an

AxioCam MRm.

Plasmid and transgenesis of epn-1::zf1::GFP CRISPR allele for epn-1 RNAi experiment

To generate epn-1(ar641[epn-1::zf1::GFP]), homology repair template pJC90 and sgRNA plasmids pJC79 and pJC80 were purified with PureLink miniprep kit, and all co- injection plasmids were purified with midi-prep (Qiagen) or ethanol precipitation. N2 animals were injected with pJC90 at 50ng/ul, pJC79 and pJC80 at 25 ng/ul each, pCFJ90 at 5 ng/ul, pGH8 at 10 ng/ul. Successful integrant was isolated and self-excised according to protocol described by Dickinson et al. (2015) The result strain has a spontaneous in-frame deletion that

removes P and I at 447 and 448aa respectively, but it does not seem to affect GFP expression when compared with other epn-1::GFP transgenes generated, or have epn-1 phenotypes, lethality or any other overt defects. epn-1 RNAi experiment

Strain GS9281 [zuIs178 [his-72::GFP]; stIs10138[tbx-38p::H1-mCherry]; epn-

1(ar641)] was used for epn-1 RNAi. Feeding RNAi was completed as described (Kamath and

Ahringer, 2003). L4 P0s were placed on RNAi plates containing HT115 bacteria expressing dsRNA specific for epn-1, lacZ (negative control) and apx-1 (positive control). Experiment was conducted at 25oC. 4-cell stage embryos were obtained by dissecting adult P0s after 24 hrs in in

10mM levamisole on a glass coverslip. Embryos were then placed on agarose pads on glass slides and were first scored for loss of epn-1::zf1::GFP. Glass slides were then placed in 25oC incubator. The same embryos with epn-1::zf1::GFP knockdown were then scored for expression of tbx-38p::H1-mCherry 2 hours later at ~150 cell stage.

EPN-1::ZF1::GFP knockdown was scored by collecting Z-stacks of GFP fluorescence at

320ms exposure time with a Zeiss spinning disk confocal dual camera system at 40x. Scoring of

150 cell stage embryos was done by collecting Z stacks of GFP (488 nm) and DsRed (561 nm) laser at 170ms and 1500ms exposure time respectively, with a Zeiss spinning disk confocal dual camera system at 40x magnification.

The number of mCherry positive nuclei and GFP nuclei were manually counted using

ImageJ with the Cell Counter plugin. Graph was created using PRISM.

Transgenesis and strains for arIs164; epn-1(0) experiment

To make a positive control strain for the experiment: arIs164[glt-3p::Venus] lag-

2(q411); arEx2511[lag-2(+)], we first generated lag-2(q420ts); arEx2511[lag-2(+)] by injection into lag-2(q420ts) with PvuI digested pJC95 at 3ng/ul, ScaI digested pCW2.1 (ceh-

22p::GFP) at 2ng/ul, and PvuII digested N2 genomic DNA at 50ng/ul. Plates were maintained at permissive temperature for 3 days, then shifted to restrictive temperature. Surviving progeny with green pharynx were isolated 7 days after injection. lag-2(q420ts); arE2511 were then used to generate arIs154;lag-2(q411); arEx2511[lag-2(+)]. Genotype of the strain is confirmed by

PCR and sequencing. Strains were maintained and experiments were performed at 25oC

Scoring of arIs164 in epn-1(0)

To score arIs164[glt-3p::Venus] in epn-1(0), 10-15 adults of epn-1(0); enEx[epn-

1::GFP] were isolated onto an OP50 plate to lay eggs for 3 hours and later removed. 24 hours later, progeny from epn-1(0); enEx[epn-1::GFP] were picked and imaged on the GFP channel at

50ms exposure time at a Zeiss Axio Imager D1 microscope with an AxioCam MRm at 40x. epn-

1(0) progeny were identified by the lack of EPN-1::GFP, which is ubiquitously expressed throughout the animal. Low exposure time was used to show presence of EPN-1::GFP, since arIs164 is very bright.

To score arIs164 in lag-2(0), lag-2(0) progeny from lag-2(0); arE2511[lag-2(+)] lacking a green pharynx were picked and imaged on the GFP channel at 700ms exposure time on a Zeiss

Axio Imager D1 microscope with an AxioCam MRm at 40x. Same was done for the control lag-

2(0); arEx2511[lag-2(+)], except images were taken at 50ms exposure time to show presence of green pharynx, since arIs164 is very bright.

Statistical Analysis

When comparing two genotypes for the frequency of two outcomes, a two-tailed 2x2

Fisher’s exact test was used. Differences were considered significant if the p-value is less than or equal to 0.05.

Key resource table Reagent or Resources Source Identifier

C. elegans: N2: C. elegans, Bristol isolate Caenorhabditis WB

Genetics Center strain: N2

C. elegans: FX30233: tmC16[unc-60(tmIS1210)]V Caenorhabditis WB

Genetics Center strain:

FX30233

C. elegans: ZH1800: unc-76(e911)V; epn-1(en47)X; enEx867 Zheng Zhou N/A

[pQS41+punc-76]

C. elegans: JCC596: unc119(ed3)III; ltIs38 [pAA1; pie- Julie Canman N/A

1/GFP::PH(PLC1delta1); unc-119 (+)] III]; zuIs178 [his-

72::SRPVAT::GFP]; stIs10024[pie-1::H2B::GFP + unc-

119(+)]. stIs10138 [tbx-38::H1-Wcherry + unc-119(+)].

C. elegans: arIs51[cdh-3p::gfp] Xantha Karp N/A

C. elegans: arIs164[glt-3p::Venus] Daniel Shaye N/A

C. elegans: GS9093: arSi48[lag-2p::lag-2(+)]; lag-2(q411)] This paper N/A

C. elegans: GS9094: arSi49[lag-2p::lag-2(KtoR)]; lag- This paper N/A

2(q411)]

C. elegans: GS9424: arSi89[lag-2p::lag-2(ΔIC)]; arIs51; lag- This paper N/A

2(q411)

C. elegans: GS9785: arSi62[lag-2p::lag-2(Random Peptideno This paper N/A

K)]; arIs51; lag-2(q411)

C. elegans: GS9786: arSi77[lag-2p::lag-2(Random Peptide)]; This paper N/A arIs51; lag-2(q411)

C. elegans: GS9398: arSi88[lag-2p::apx-1(+)]; lag-2(q411) This paper N/A

C. elegans: GS9439: arSi90[lag-2p::apx-1(KtoR)]; arIs51; lag- This paper N/A

2(q411)

C. elegans:GS9440: arSi91[lag-2p::apx-1(ΔIC)]; arIs51; lag- This paper N/A

2(q411)

C. elegans: GS9216: epn-1(ar641 [epn-1::zf1::GFP]) This paper N/A

C. elegans: GS9281: zuIs178 [his-72::GFP]; stIs10138[tbx- This paper N/A

38::H1-mCherry]; epn-1(ar641 [epn-1::zf1::GFP])

C. elegans: GS9396: arIs164; epn-1(en47); enEx[epn-1::GFP] This paper N/A

C. elegans: GS9397: arIs164 lag-2(q411); arEx[lag-2(+)] This paper N/A

C. elegans: GS9372: lag-2(q411); arEx2523[apx-1(secreted)] This paper N/A

C. elegans: GS9373: lag-2(q411); arEx[apx-1(secreted)] This paper N/A

C. elegans: GS9374: lag-2(q411); arEx[apx-1(secreted)] This paper N/A

C. elegans: GS9418: arIs51; lag-2(q411); arEx2523[lag- This paper N/A

2p::apx-1(secreted)]

C. elegans: GS9419: arIs51; lag-2(q411); arEx2530[lag- This paper N/A

2p::apx-1(secreted)]

C. elegans: GS9420: arIs51; lag-2(q411); arEx2531[lag- This paper N/A

2p::apx-1(secreted)]

C. elegans: GS9412: lag-2(q411)/tmC16; arEx2524[lag- This paper N/A

2p::lag-2(secreted)]

C. elegans: GS0413: lag-2(q411)/tmC16; arEx2525[lag- This paper N/A

2p::lag-2(secreted)]

C. elegans: GS9414: lag-2(q411)/tmC16; arEx2526[lag- This paper N/A

2p::lag-2(secreted)]

C. elegans: GS9415: lag-2(q411)/tmC16; arEx2527[lag- This paper N/A

2p::dsl-1(secreted)]

C. elegans: GS9416: lag-2(q411)/tmC16; arEx2528[lag- This paper N/A

2p::dsl-1(secreted)]

C. elegans: GS9417: lag-2(q411)/tmC16; arEx2529[lag- This paper N/A

2p::dsl-1(secreted)]

C. elegans: GS9510: lag-2(q411); arEx2533[lag-2p::lag- This paper N/A

2(secreted)::GPI]

C. elegans: GS9511: lag-2(q411); arEx2534[lag-2p::lag- This paper N/A

2(secreted)::GPI]

C. elegans: GS9512: lag-2(q411); arEx2535[lag-2p::lag- This paper N/A

2(secreted)::GPI]

C. elegans: GS9513: lag-2(q411); arEx2536[lag-2p::lag- This paper N/A

2(secreted)]

C. elegans: GS9514: lag-2(q411); arEx2537[lag-2p::lag- This paper N/A

2(secreted)]

C. elegans: GS9515: lag-2(q411); arEx2538[lag-2p::lag- This paper N/A

2(secreted)]

C. elegans: GS9526: lag-2(q411); tmC16 This paper N/A

C. elegans: GS9527: lag-2(q411); arSi95[lag-2p::lag- This paper N/A

2(secreted)-GPI]

Recombinant DNA Reagents and resources Source Identifier

Plasmid pCFJ90 Addgene #19327

Plasmid pGH8 Addgene #19359

Plasmid pWZ111 Bob Goldstein N/A

Plasmid pAP082 Bob Goldstein N/A

Plasmid pQS41 Zheng Zhou N/A

Plasmid pJC70 lag-2p(3.3kb)::lag-2(+) This paper N/A

Plasmid pJC71 lag-2p(3.3kb)::lag-2(KtoR) This paper N/A

Plasmid pJC73 lag-2p(3.3kb)::lag-2(Random Peptide no K) This paper N/A

Plasmid pJC91 lag-2p(3.3kb)::lag-2(Random Peptide) This paper N/A

Plasmid pJC99 lag-2p(3.3kb)::lag-2(secreted) This paper N/A

Plasmid pJC106 lag-2p(3.3kb)::lag-2(ΔIC) This paper N/A

Plasmid pJC100 lag-2p(3.3kb)::apx-1 cDNA(+)::lag-2 This paper N/A

3’UTR (1kb)

Plasmid pJC120 lag-2p(3.3kb)::apx-1 cDNA(K to R)::lag-2 This paper N/A

3’UTR (1kb)

Plasmid pJC121 lag-2p(3.3kb)::apx-1 cDNA(ΔIC)::lag-2 This paper N/A

3’UTR (1kb)

Plasmid pJC109 lag-2p(3.3kb)::apx-1 cDNA(secreted)::lag- This paper N/A

2 3’UTR (1kb)

Plasmid pJC79 epn-1(ar641) 5’ sgRNA This paper N/A

Plasmid pJC80 epn-1(ar641) 3’ sgRNA This paper N/A

Plasmid pJC90 epn-1(ar641) homology repair template This paper N/A

Plasmid pJC119 lag-2p(3.3kb)::lag-2(secreted)::GPI This paper N/A

Plasmid pJC115 lag-2p(3.3kb):dsl-1cDNA This paper N/A

References

Attner, M.A., Keil, W., Benavidez, J.M., and Greenwald, I. (2019). HLH-2/E2A Expression Links Stochastic and Deterministic Elements of a Cell Fate Decision during C. elegans Gonadogenesis. Current biology : CB 29, 3094-3100.e3094. Benhra, N., Vignaux, F., Dussert, A., Schweisguth, F., and Le Borgne, R. (2010). Neuralized promotes basal to apical transcytosis of delta in epithelial cells. Molecular biology of the cell 21, 2078-2086. Brenner, J.L., and Schedl, T. (2016). Germline Stem Cell Differentiation Entails Regional Control of Cell Fate Regulator GLD-1 in Caenorhabditis elegans. Genetics 202, 1085-1103. Brou, C., Logeat, F., Gupta, N., Bessia, C., LeBail, O., Doedens, J.R., Cumano, A., Roux, P., Black, R.A., and Israel, A. (2000). A novel proteolytic cleavage involved in Notch signaling: the role of the disintegrin-metalloprotease TACE. Molecular cell 5, 207-216. Byrd, D.T., Knobel, K., Affeldt, K., Crittenden, S.L., and Kimble, J. (2014). A DTC niche plexus surrounds the germline stem cell pool in Caenorhabditis elegans. PloS one 9, e88372. Chen, H., Ko, G., Zatti, A., Di Giacomo, G., Liu, L., Raiteri, E., Perucco, E., Collesi, C., Min, W., Zeiss, C., et al. (2009). Embryonic arrest at midgestation and disruption of Notch signaling produced by the absence of both epsin 1 and epsin 2 in mice. Proceedings of the National Academy of Sciences of the United States of America 106, 13838-13843. Chen, N., and Greenwald, I. (2004). The lateral signal for LIN-12/Notch in C. elegans vulval development comprises redundant secreted and transmembrane DSL proteins. Developmental cell 6, 183-192. Choi, M.S., Yoo, A.S., and Greenwald, I. (2010). sel-11 and cdc-42, two negative modulators of LIN-12/Notch activity in C. elegans. PloS one 5, e11885. Christensen, S., Kodoyianni, V., Bosenberg, M., Friedman, L., and Kimble, J. (1996). lag-1, a gene required for lin-12 and glp-1 signaling in Caenorhabditis elegans, is homologous to human CBF1 and Drosophila Su(H). Development (Cambridge, England) 122, 1373-1383. Clark, S.G., Shurland, D.L., Meyerowitz, E.M., Bargmann, C.I., and van der Bliek, A.M. (1997). A dynamin GTPase mutation causes a rapid and reversible temperature-inducible locomotion defect in C. elegans. Proceedings of the National Academy of Sciences of the United States of America 94, 10438-10443. Cordle, J., Johnson, S., Tay, J.Z., Roversi, P., Wilkin, M.B., de Madrid, B.H., Shimizu, H., Jensen, S., Whiteman, P., Jin, B., et al. (2008). A conserved face of the Jagged/Serrate DSL domain is involved in Notch trans-activation and cis-inhibition. Nature structural & molecular biology 15, 849-857.

Crittenden, S.L., Bernstein, D.S., Bachorik, J.L., Thompson, B.E., Gallegos, M., Petcherski, A.G., Moulder, G., Barstead, R., Wickens, R., Kimble, J. (2002) A conserved RNA-binding protein controls germline stem cells in Caenorhabditis elegans. Nature 417, 660-3. Crittenden, S.L., Lee, C., Mohanty, I., Battula, S., Knobel, K., and Kimble, J. (2019). Sexual dimorphism of niche architecture and regulation of the Caenorhabditis elegans germline stem cell pool. Molecular biology of the cell 30, 1757-1769. de Celis, J.F., and Bray, S. (1997). Feed-back mechanisms affecting Notch activation at the dorsoventral boundary in the Drosophila wing. Development (Cambridge, England) 124, 3241- 3251. Doyle, T.G., Wen, C., and Greenwald, I. (2000). SEL-8, a nuclear protein required for LIN-12 and GLP-1 signaling in Caenorhabditis elegans. Proceedings of the National Academy of Sciences of the United States of America 97, 7877-7881. Eisenhaber, B., Bork, P., and Eisenhaber, F. (1999). Prediction of potential GPI-modification sites in proprotein sequences. Journal of molecular biology 292, 741-758. Emery, G., Hutterer, A., Berdnik, D., Mayer, B., Wirtz-Peitz, F., Gaitan, M.G., and Knoblich, J.A. (2005). Asymmetric Rab 11 endosomes regulate delta recycling and specify cell fate in the Drosophila nervous system. Cell 122, 763-773. Fankhauser, N., and Maser, P. (2005). Identification of GPI anchor attachment signals by a Kohonen self-organizing map. Bioinformatics (Oxford, England) 21, 1846-1852. Fitzgerald, K., and Greenwald, I. (1995). Interchangeability of Caenorhabditis elegans DSL proteins and intrinsic signalling activity of their extracellular domains in vivo. Development (Cambridge, England) 121, 4275-4282. Forrester, W.C., Garriga, G. (1997) Genes necessary for C. elegans cell and growth cone migrations. Development 124, 1831-43. Gazave, E., Lapebie, P., Richards, G.S., Brunet, F., Ereskovsky, A.V., Degnan, B.M., Borchiellini, C., Vervoort, M., and Renard, E. (2009). Origin and evolution of the Notch signalling pathway: an overview from eukaryotic genomes. BMC evolutionary biology 9, 249. Gíslason, M.H., Nielsen, H., Almagro Armenteros, J.J., and Johansen, A.R. (2019). Prediction of GPI-Anchored proteins with pointer neural networks. 838680. Good, K., Ciosk, R., Nance, J., Neves, A., Hill, R.J., and Priess, J.R. (2004). The T-box transcription factors TBX-37 and TBX-38 link GLP-1/Notch signaling to mesoderm induction in C. elegans embryos. Development (Cambridge, England) 131, 1967-1978. Goode, S., Melnick, M., Chou, T.B., and Perrimon, N. (1996). The neurogenic genes egghead and brainiac define a novel signaling pathway essential for epithelial morphogenesis during Drosophila oogenesis. Development (Cambridge, England) 122, 3863-3879.

Gordon, W.R., Roy, M., Vardar-Ulu, D., Garfinkel, M., Mansour, M.R., Aster, J.C., and Blacklow, S.C. (2009). Structure of the Notch1-negative regulatory region: implications for normal activation and pathogenic signaling in T-ALL. Blood 113, 4381-4390. Gordon, W.R., Vardar-Ulu, D., Histen, G., Sanchez-Irizarry, C., Aster, J.C., and Blacklow, S.C. (2007). Structural basis for autoinhibition of Notch. Nature structural & molecular biology 14, 295-300. Gordon, W.R., Zimmerman, B., He, L., Miles, L.J., Huang, J., Tiyanont, K., McArthur, D.G., Aster, J.C., Perrimon, N., Loparo, J.J., et al. (2015). Mechanical Allostery: Evidence for a Force Requirement in the Proteolytic Activation of Notch. Developmental cell 33, 729-736. Greenwald, I.S., Sternberg, P.W., and Horvitz, H.R. (1983). The lin-12 locus specifies cell fates in Caenorhabditis elegans. Cell 34, 435-444. Hansen, D., Shedl, T. (2013) Stem cell proliferation versus meiotic fate decision in Caenorhbditis elegans. Adv Exp Med Biol 757, 71-99. Hansen, D., Wilson-Berry, L., Dang, T., and Schedl, T. (2004). Control of the proliferation versus meiotic development decision in the C. elegans germline through regulation of GLD-1 protein accumulation. 131, 93-104. Hansen, D., Wilson-Berry, L., Dang, T., Schedl, T. (2004) Control of the proliferation versus meiotic development decision in the C. elegans germline through regulation of GLD-1 protein accumulation. Henderson, S.T., Gao, D., Christensen, S., and Kimble, J. (1997). Functional domains of LAG-2, a putative signaling ligand for LIN-12 and GLP-1 receptors in Caenorhabditis elegans. Molecular biology of the cell 8, 1751-1762. Henderson, S.T., Gao, D., Lambie, E.J., and Kimble, J. (1994). lag-2 may encode a signaling ligand for the GLP-1 and LIN-12 receptors of C. elegans. Development (Cambridge, England) 120, 2913-2924. Herrera, L.A., and Starr, D.A. (2018). The E3 Ubiquitin Ligase MIB-1 Is Necessary To Form the Nuclear Halo in Caenorhabditis elegans Sperm. G3 (Bethesda, Md) 8, 2465-2470. Hill, R.J., and Sternberg, P.W. (1992). The gene lin-3 encodes an inductive signal for vulval development in C. elegans. Nature 358, 470-476. Hirsch, S.M., Sundaramoorthy, S., Davies, T., Zhuravlev, Y., Waters, J.C., Shirasu-Hiza, M., Dumont, J., and Canman, J.C. (2018). FLIRT: fast local infrared thermogenetics for subcellular control of protein function. Nature methods 15, 921-923. Hukriede, N.A., Gu, Y., and Fleming, R.J. (1997). A dominant-negative form of Serrate acts as a general antagonist of Notch activation. Development (Cambridge, England) 124, 3427-3437.

Hutter, H., and Schnabel, R. (1995). Establishment of left-right asymmetry in the Caenorhabditis elegans embryo: a multistep process involving a series of inductive events. Development (Cambridge, England) 121, 3417-3424. Hutter, H., and Schnabel, R. (1995). Specification of anterior-posterior differences within the AB lineage in the C. elegans embryo: a polarising induction. Development (Cambridge, England) 121, 1559-1568. Itoh, M., Kim, C.H., Palardy, G., Oda, T., Jiang, Y.J., Maust, D., Yeo, S.Y., Lorick, K., Wright, G.J., Ariza-McNaughton, L., et al. (2003). Mind bomb is a ubiquitin ligase that is essential for efficient activation of Notch signaling by Delta. Developmental cell 4, 67-82. Jacobsen, T.L., Brennan, K., Arias, A.M., and Muskavitch, M.A. (1998). Cis-interactions between Delta and Notch modulate neurogenic signalling in Drosophila. Development (Cambridge, England) 125, 4531-4540. Jafar-Nejad, H., Andrews, H.K., Acar, M., Bayat, V., Wirtz-Peitz, F., Mehta, S.Q., Knoblich, J.A., and Bellen, H.J. (2005). Sec15, a component of the exocyst, promotes notch signaling during the asymmetric division of Drosophila sensory organ precursors. Developmental cell 9, 351-363. Jarriault, S., and Greenwald, I. (2005). Evidence for functional redundancy between C. elegans ADAM proteins SUP-17/Kuzbanian and ADM-4/TACE. Developmental biology 287, 1-10. Kang, Y.L., Yochem, J., Bell, L., Sorensen, E.B., Chen, L., and Conner, S.D. (2013). Caenorhabditis elegans reveals a FxNPxY-independent low-density lipoprotein receptor internalization mechanism mediated by epsin1. Molecular biology of the cell 24, 308-318. Kasbauer, T., Towb, P., Alexandrova, O., David, C.N., Dall'armi, E., Staudigl, A., Stiening, B., and Bottger, A. (2007). The Notch signaling pathway in the cnidarian Hydra. Developmental biology 303, 376-390. Katic, I., Vallier, L.G., and Greenwald, I. (2005). New positive regulators of lin-12 activity in Caenorhabditis elegans include the BRE-5/Brainiac glycosphingolipid biosynthesis enzyme. Genetics 171, 1605-1615. Kershner, A., Crittenden, S.L., Friend, K., Sorensen, E.B., Porter, D.F., Kimble, J. (2013) Germline stem cells and their regulation in the nematode Caenorhabditis elegans. Adv Exp Med Biol 786, 29-46. Kimble, J. (1981). Alterations in cell lineage following laser ablation of cells in the somatic gonad of Caenorhabditis elegans. Developmental biology 87, 286-300. Kimble, J., Hirsch, D. (1979) The postembryonic cell lineages of the hermaphrodite and male gonads in Caenorhabditis elegans. Dev Biol 70, 396-417. Klein, T., and Arias, A.M. (1998). Interactions among Delta, Serrate and Fringe modulate Notch activity during Drosophila wing development. Development (Cambridge, England) 125, 2951- 2962.

Klein, T., Brennan, K., and Arias, A.M. (1997). An intrinsic dominant negative activity of serrate that is modulated during wing development in Drosophila. Developmental biology 189, 123-134. Komatsu, H., Chao, M.Y., Larkins-Ford, J., Corkins, M.E., Somers, G.A., Tucey, T., Dionne, H.M., White, J.Q., Wani, K., Boxem, M., et al. (2008). OSM-11 facilitates LIN-12 Notch signaling during Caenorhabditis elegans vulval development. PLoS biology 6, e196. Kwon, D.Y., Dimitriadi, M., Terzic, B., Cable, C., Hart, A.C., Chitnis, A., Fischbeck, K.H., and Burnett, B.G. (2013). The E3 ubiquitin ligase mind bomb 1 ubiquitinates and promotes the degradation of survival of motor neuron protein. Molecular biology of the cell 24, 1863-1871. Lambie, E.J., Kimble, J.(1991) Two homologous regulatory genes, lin-12 and glp-1, have overlapping functions. Development 112, 231-40. Langridge, P.D., and Struhl, G. (2017). Epsin-Dependent Ligand Endocytosis Activates Notch by Force. Cell 171, 1383-1396.e1312. Levitan, D., and Greenwald, I. (1995). Facilitation of lin-12-mediated signaling by sel-12, a Caenorhabditis elegans S182 Alzheimer's disease gene. Nature 377, 351-354. Li, X., and Greenwald, I. (1997). HOP-1, a Caenorhabditis elegans presenilin, appears to be functionally redundant with SEL-12 presenilin and to facilitate LIN-12 and GLP-1 signaling. Proceedings of the National Academy of Sciences of the United States of America 94, 12204- 12209. lineage of the nematode Caenorhabditis elegans. Dev Biol 100, 64–119. Liu, Y., and Maine, E.M. (2007). The Bro1-domain protein, EGO-2, promotes Notch signaling in Caenorhabditis elegans. Genetics 176, 2265-2277. Logeat, F., Bessia, C., Brou, C., LeBail, O., Jarriault, S., Seidah, N.G., and Israel, A. (1998). The Notch1 receptor is cleaved constitutively by a furin-like convertase. Proceedings of the National Academy of Sciences of the United States of America 95, 8108-8112. Luca, V.C., Kim, B.C., Ge, C., Kakuda, S., Wu, D., Roein-Peikar, M., Haltiwanger, R.S., Zhu, C., Ha, T., and Garcia, K.C. (2017). Notch-Jagged complex structure implicates a catch bond in tuning ligand sensitivity. Science (New York, NY) 355, 1320-1324. Mango, S.E., Lambir, E.J., Kimble, J. (1994) The pha-4 gene is required to generate the pharyngeal primordium of Caenorhabditis elegans. Development 120, 3019-31 Marlow, H., Roettinger, E., Boekhout, M., and Martindale, M.Q. (2012). Functional roles of Notch signaling in the cnidarian Nematostella vectensis. Developmental biology 362, 295-308. Marsh, M., and Helenius, A. (2006). Virus entry: open sesame. Cell 124, 729-740. Mayor, S., and Pagano, R.E. (2007). Pathways of clathrin-independent endocytosis. Nature reviews Molecular cell biology 8, 603-612.

McDaniell, R., Warthen, D.M., Sanchez-Lara, P.A., Pai, A., Krantz, I.D., Piccoli, D.A., and Spinner, N.B. (2006). NOTCH2 mutations cause Alagille syndrome, a heterogeneous disorder of the notch signaling pathway. American journal of human genetics 79, 169-173. McGovern, M., Voutev, R., Maciejowski, J., Corsi, A.K., and Hubbard, E.J. (2009). A "latent niche" mechanism for tumor initiation. Proceedings of the National Academy of Sciences of the United States of America 106, 11617-11622. Mello, C.C., Draper, B.W., Priess, J.R. (1994). The maternal genes apx-1 and glp-1 and establishment of dorsal-ventral polarity in the early C. elegans embryo. Cell 77,95-106 Meloty-Kapella, L., Shergill, B., Kuon, J., Botvinick, E., and Weinmaster, G. (2012). Notch ligand endocytosis generates mechanical pulling force dependent on dynamin, epsins, and actin. Developmental cell 22, 1299-1312. Mickey, K.M., Mello, CC., Montogomery, M.K., Fire, A., Priess J.R. (1996) An inductive interaction in 4-cell stage C. elegans embryos involves APX-1 expression in the signaling cell. Development 122, 1791-1798 Moloney, D.J., Panin, V.M., Johnston, S.H., Chen, J., Shao, L., Wilson, R., Wang, Y., Stanley, P., Irvine, K.D., Haltiwanger, R.S., et al. (2000). Fringe is a glycosyltransferase that modifies Notch. Nature 406, 369-375. Morrissette, J.D., Colliton, R.P., and Spinner, N.B. (2001). Defective intracellular transport and processing of JAG1 missense mutations in Alagille syndrome. Human molecular genetics 10, 405-413. Moskowitz, I.P., and Rothman, J.H. (1996). lin-12 and glp-1 are required zygotically for early embryonic cellular interactions and are regulated by maternal GLP-1 signaling in Caenorhabditis elegans. Development (Cambridge, England) 122, 4105-4117. Munder, S., Tischer, S., Grundhuber, M., Buchels, N., Bruckmeier, N., Eckert, S., Seefeldt, C.A., Prexl, A., Kasbauer, T., and Bottger, A. (2013). Notch-signalling is required for head regeneration and tentacle patterning in Hydra. Developmental biology 383, 146-157. Nakayama, Y., Shivas, J.M., Poole, D.S., Squirrell, J.M., Kulkoski, J.M., Schleede, J.B., and Skop, A.R. (2009). Dynamin participates in the maintenance of anterior polarity in the Caenorhabditis elegans embryo. Developmental cell 16, 889-900. Nandagopal, N., Santat, L.A., LeBon, L., Sprinzak, D., Bronner, M.E., and Elowitz, M.B. (2018). Dynamic Ligand Discrimination in the Notch Signaling Pathway. Cell 172, 869- 880.e819. Nelson, F.K., Riddle, D.L. (1984) Functional study of the Caenorhabditis elegans secretory- excretory system using laser microsurgery. J Exp Zool 231, 45-56 Neves, A., and Priess, J.R. (2005). The REF-1 family of bHLH transcription factors pattern C. elegans embryos through Notch-dependent and Notch-independent pathways. Developmental cell 8, 867-879.

Parker, S., Peterkin, H.S., and Baylis, H.A. (2007). Muscular dystrophy associated mutations in caveolin-1 induce neurotransmission and locomotion defects in Caenorhabditis elegans. Invertebrate neuroscience : IN 7, 157-164. Parker, S., Walker, D.S., Ly, S., and Baylis, H.A. (2009). Caveolin-2 is required for apical lipid trafficking and suppresses basolateral recycling defects in the intestine of Caenorhabditis elegans. Molecular biology of the cell 20, 1763-1771. Parks, A.L., Klueg, K.M., Stout, J.R., and Muskavitch, M.A. (2000). Ligand endocytosis drives receptor dissociation and activation in the Notch pathway. Development (Cambridge, England) 127, 1373-1385. Parks, A.L., Stout, J.R., Shepard, S.B., Klueg, K.M., Dos Santos, A.A., Parody, T.R., Vaskova, M., and Muskavitch, M.A. (2006). Structure-function analysis of delta trafficking, receptor binding and signaling in Drosophila. Genetics 174, 1947-1961. Parton, R.G., and Simons, K. (2007). The multiple faces of caveolae. Nature reviews Molecular cell biology 8, 185-194. Petcherski, A.G., and Kimble, J. (2000). Mastermind is a putative activator for Notch. Current biology : CB 10, R471-473. Pierleoni, A., Martelli, P.L., and Casadio, R. (2008). PredGPI: a GPI-anchor predictor. BMC bioinformatics 9, 392. Powell, P.A., Wesley, C., Spencer, S., and Cagan, R.L. (2001). Scabrous complexes with Notch to mediate boundary formation. Nature 409, 626-630. Priess, J.R. (2005). Notch signaling in the C. elegans embryo. WormBook : the online review of C elegans biology, 1-16. Rajan, A., Tien, A.C., Haueter, C.M., Schulze, K.L., and Bellen, H.J. (2009). The Arp2/3 complex and WASp are required for apical trafficking of Delta into microvilli during cell fate specification of sensory organ precursors. Nature cell biology 11, 815-824. Ratliff, M., Hill-Harfe, K.L., Gleason, E.J., Ling, H., Kroft, T.L., and L'Hernault, S.W. (2018). MIB-1 Is Required for Spermatogenesis and Facilitates LIN-12 and GLP-1 Activity in Caenorhabditis elegans. Genetics 209, 173-193. Richards, G.S., and Degnan, B.M. (2012). The expression of Delta ligands in the sponge Amphimedon queenslandica suggests an ancient role for Notch signaling in metazoan development. EvoDevo 3, 15. Richards, G.S., and Degnan, B.M. (2012). The expression of Delta ligands in the sponge Amphimedon queenslandica suggests an ancient role for Notch signaling in metazoan development. EvoDevo 3, 15.

Richards, G.S., Simionato, E., Perron, M., Adamska, M., Vervoort, M., and Degnan, B.M. (2008). Sponge genes provide new insight into the evolutionary origin of the neurogenic circuit. Current biology : CB 18, 1156-1161. Ringrose, J.H., van den Toorn, H.W., Eitel, M., Post, H., Neerincx, P., Schierwater, B., Altelaar, A.F., and Heck, A.J. (2013). Deep proteome profiling of Trichoplax adhaerens reveals remarkable features at the origin of metazoan multicellularity. Nature communications 4, 1408. Sakata, T., Sakaguchi, H., Tsuda, L., Higashitani, A., Aigaki, T., Matsuno, K., and Hayashi, S. (2004). Drosophila Nedd4 regulates endocytosis of notch and suppresses its ligand-independent activation. Current biology : CB 14, 2228-2236. Sanchez-Irizarry, C., Carpenter, A.C., Weng, A.P., Pear, W.S., Aster, J.C., and Blacklow, S.C. (2004). Notch subunit heterodimerization and prevention of ligand-independent proteolytic activation depend, respectively, on a novel domain and the LNR repeats. Molecular and cellular biology 24, 9265-9273. Sato, K., Ernstrom, G.G., Watanabe, S., Weimer, R.M., Chen, C.H., Sato, M., Siddiqui, A., Jorgensen, E.M., and Grant, B.D. (2009). Differential requirements for clathrin in receptor- mediated endocytosis and maintenance of synaptic vesicle pools. Proceedings of the National Academy of Sciences of the United States of America 106, 1139-1144. Scheel, J., Srinivasan, J., Honnert, U., Henske, A., and Kurzchalia, T.V. (1999). Involvement of caveolin-1 in meiotic cell-cycle progression in Caenorhabditis elegans. Nature cell biology 1, 127-129. Seo, D., Southard, K.M., Kim, J.W., Lee, H.J., Farlow, J., Lee, J.U., Litt, D.B., Haas, T., Alivisatos, A.P., Cheon, J., et al. (2016). A Mechanogenetic Toolkit for Interrogating Cell Signaling in Space and Time. Cell 165, 1507-1518. Seugnet, L., Simpson, P., and Haenlin, M. (1997). Requirement for dynamin during Notch signaling in Drosophila neurogenesis. Developmental biology 192, 585-598. Seydoux, G., and Greenwald, I. (1989). Cell autonomy of lin-12 function in a cell fate decision in C. elegans. Cell 57, 1237-1245. Shen, Q., He, B., Lu, N., Conradt, B., Grant, B.D., and Zhou, Z. (2013). Phagocytic receptor signaling regulates clathrin and epsin-mediated cytoskeletal remodeling during apoptotic cell engulfment in C. elegans. Development (Cambridge, England) 140, 3230-3243. Shimizu, K., Chiba, S., Kumano, K., Hosoya, N., Takahashi, T., Kanda, Y., Hamada, Y., Yazaki, Y., and Hirai, H. (1999). Mouse jagged1 physically interacts with notch2 and other notch receptors. Assessment by quantitative methods. The Journal of biological chemistry 274, 32961- 32969. Sorensen, E.B., and Conner, S.D. (2010). gamma-secretase-dependent cleavage initiates notch signaling from the plasma membrane. Traffic (Copenhagen, Denmark) 11, 1234-1245.

Srivastava, M., Simakov, O., Chapman, J., Fahey, B., Gauthier, M.E.A., Mitros, T., Richards, G.S., Conaco, C., Dacre, M., Hellsten, U., et al. (2010). The Amphimedon queenslandica genome and the evolution of animal complexity. Nature 466, 720-726. Stephenson, N.L., and Avis, J.M. (2012). Direct observation of proteolytic cleavage at the S2 site upon forced unfolding of the Notch negative regulatory region. Proceedings of the National Academy of Sciences of the United States of America 109, E2757-2765. Sternberg, P.W. (2005). Vulval development. WormBook : the online review of C elegans biology, 1-28. Struhl, G., and Greenwald, I. (1999). Presenilin is required for activity and nuclear access of Notch in Drosophila. Nature 398, 522-525. Struhl, G., and Greenwald, I. (2001). Presenilin-mediated transmembrane cleavage is required for Notch signal transduction in Drosophila. Proceedings of the National Academy of Sciences of the United States of America 98, 229-234. Struhl, G., Fitzgerald, K., and Greenwald, I. (1993). Intrinsic activity of the Lin-12 and Notch intracellular domains in vivo. Cell 74, 331-345. Suh, N., Jedamzik, B., Eckmann, C.R., Wickens, M., and Kimble, J. (2006). The GLD-2 poly(A) polymerase activates gld-1 mRNA in the Caenorhabditis elegans germ line. Proceedings of the National Academy of Sciences of the United States of America 103, 15108-15112. Sulston, J.E., Schierenberg, E., White, J.G., and Thomson, J.N. (1983). The embryonic cell Sun, X., and Artavanis-Tsakonas, S. (1996). The intracellular deletions of Delta and Serrate define dominant negative forms of the Drosophila Notch ligands. Development (Cambridge, England) 122, 2465-2474. Sun, X., and Artavanis-Tsakonas, S. (1997). Secreted forms of DELTA and SERRATE define antagonists of Notch signaling in Drosophila. Development (Cambridge, England) 124, 3439- 3448. Sundaram, M.V., Buechner, M. (2016) The Caenorhabditis elegans excretory system: a model for tubulogenesis, cell fate specification, and plasticity. Genetics 203, 35-63. Tax, F.E., Yeargers, J.J., and Thomas, J.H. (1994). Sequence of C. elegans lag-2 reveals a cell- signalling domain shared with Delta and Serrate of Drosophila. Nature 368, 150-154. Tian, X., Hansen, D., Schedl, T., and Skeath, J.B. (2004). Epsin potentiates Notch pathway activity in Drosophila and C. elegans. Development (Cambridge, England) 131, 5807-5815. Vooijs, M., Schroeter, E.H., Pan, Y., Blandford, M., and Kopan, R. (2004). Ectodomain shedding and intramembrane cleavage of mammalian Notch proteins is not regulated through oligomerization. The Journal of biological chemistry 279, 50864-50873. Wang, W., and Struhl, G. (2004). Drosophila Epsin mediates a select endocytic pathway that DSL ligands must enter to activate Notch. Development (Cambridge, England) 131, 5367-5380.

Wang, W., and Struhl, G. (2005). Distinct roles for Mind bomb, Neuralized and Epsin in mediating DSL endocytosis and signaling in Drosophila. Development (Cambridge, England) 132, 2883-2894. Wang, X., Roy, P.J., Holland, S.J., Zhang, L.W., Culotti, J.G., and Pawson, T. (1999). Multiple ephrins control cell organization in C. elegans using kinase-dependent and -independent functions of the VAB-1 Eph receptor. Molecular cell 4, 903-913. Wesley, C.S. (1999). Notch and wingless regulate expression of cuticle patterning genes. Molecular and cellular biology 19, 5743-5758. Wilkinson, H.A., Fitzgerald, K., and Greenwald, I. (1994). Reciprocal changes in expression of the receptor lin-12 and its ligand lag-2 prior to commitment in a C. elegans cell fate decision. Cell 79, 1187-1198. Wu, L., Aster, J.C., Blacklow, S.C., Lake, R., Artavanis-Tsakonas, S., and Griffin, J.D. (2000). MAML1, a human homologue of Drosophila mastermind, is a transcriptional co-activator for NOTCH receptors. Nature genetics 26, 484-489. Xie, X., Cho, B., and Fischer, J.A. (2012). Drosophila Epsin's role in Notch ligand cells requires three Epsin protein functions: the lipid binding function of the ENTH domain, a single Ubiquitin interaction motif, and a subset of the C-terminal protein binding modules. Developmental biology 363, 399-412. Xu, A.J., and Springer, T.A. (2013). Mechanisms by which von Willebrand disease mutations destabilize the A2 domain. The Journal of biological chemistry 288, 6317-6324. Yu, X., Odera, S., Chuang, C.H., Lu, N., and Zhou, Z. (2006). C. elegans Dynamin mediates the signaling of phagocytic receptor CED-1 for the engulfment and degradation of apoptotic cells. Developmental cell 10, 743-757. Zhang, W., Hayashizaki, Y., and Kone, B.C. (2004). Structure and regulation of the mDot1 gene, a mouse histone H3 methyltransferase. The Biochemical journal 377, 641-651. Zhang, X., Halvorsen, K., Zhang, C.Z., Wong, W.P., and Springer, T.A. (2009). Mechanoenzymatic cleavage of the ultralarge vascular protein von Willebrand factor. Science (New York, NY) 324, 1330-1334. Aguirre, A., Rubio, M.E., and Gallo, V. (2010). Notch and EGFR pathway interaction regulates neural stem cell number and self-renewal. Nature 467, 323-327. Berset, T., Hoier, E.F., Battu, G., Canevascini, S., and Hajnal, A. (2001). Notch inhibition of RAS signaling through MAP kinase phosphatase LIP-1 during C. elegans vulval development. Science (New York, NY) 291, 1055-1058. Bonifacino, J.S., and Traub, L.M. (2003). Signals for sorting of transmembrane proteins to endosomes and lysosomes. Annual review of biochemistry 72, 395-447.

Chang, C., Hopper, N.A., and Sternberg, P.W. (2000). Caenorhabditis elegans SOS-1 is necessary for multiple RAS-mediated developmental signals. The EMBO journal 19, 3283-3294. Clark, S.G., Chisholm, A.D., and Horvitz, H.R. (1993). Control of cell fates in the central body region of C. elegans by the homeobox gene lin-39. Cell 74, 43-55. Clark, S.G., Stern, M.J., and Horvitz, H.R. (1992). C. elegans cell-signalling gene sem-5 encodes a protein with SH2 and SH3 domains. Nature 356, 340-344. Deng, Y., and Greenwald, I. (2016). Determinants in the LIN-12/Notch Intracellular Domain That Govern Its Activity and Stability During Caenorhabditis elegans Vulval Development. G3 (Bethesda, Md) 6, 3663-3670. Dickinson, D.J., Pani, A.M., Heppert, J.K., Higgins, C.D., and Goldstein, B. (2015). Streamlined Genome Engineering with a Self-Excising Drug Selection Cassette. Genetics 200, 1035-1049. Fanto, M., and Mlodzik, M. (1999). Asymmetric Notch activation specifies photoreceptors R3 and R4 and planar polarity in the Drosophila eye. Nature 397, 523-526. Fire, A., and Waterston, R.H. (1989). Proper expression of myosin genes in transgenic nematodes. The EMBO journal 8, 3419-3428. Frokjaer-Jensen, C., Davis, M.W., Sarov, M., Taylor, J., Flibotte, S., LaBella, M., Pozniakovsky, A., Moerman, D.G., and Jorgensen, E.M. (2014). Random and targeted transgene insertion in Caenorhabditis elegans using a modified Mos1 transposon. Nature methods 11, 529-534. Gleason, J.E., Szyleyko, E.A., and Eisenmann, D.M. (2006). Multiple redundant Wnt signaling components function in two processes during C. elegans vulval development. Developmental biology 298, 442-457. Guo, S., Liu, M., and Gonzalez-Perez, R.R. (2011). Role of Notch and its oncogenic signaling crosstalk in breast cancer. Biochimica et biophysica acta 1815, 197-213. Han, M., Golden, A., Han, Y., and Sternberg, P.W. (1993). C. elegans lin-45 raf gene participates in let-60 ras-stimulated vulval differentiation. Nature 363, 133-140. Hitoshi, S., Alexson, T., Tropepe, V., Donoviel, D., Elia, A.J., Nye, J.S., Conlon, R.A., Mak, T.W., Bernstein, A., and van der Kooy, D. (2002). Notch pathway molecules are essential for the maintenance, but not the generation, of mammalian neural stem cells. Genes & development 16, 846-858. Hubbard, E.J., Wu, G., Kitajewski, J., and Greenwald, I. (1997). sel-10, a negative regulator of lin-12 activity in Caenorhabditis elegans, encodes a member of the CDC4 family of proteins. Genes & development 11, 3182-3193. Jacobs, D., Beitel, G.J., Clark, S.G., Horvitz, H.R., and Kornfeld, K. (1998). Gain-of-function mutations in the Caenorhabditis elegans lin-1 ETS gene identify a C-terminal regulatory domain phosphorylated by ERK MAP kinase. Genetics 149, 1809-1822.

Kaech, S.M., Whitfield, C.W., and Kim, S.K. (1998). The LIN-2/LIN-7/LIN-10 complex mediates basolateral membrane localization of the C. elegans EGF receptor LET-23 in vulval epithelial cells. Cell 94, 761-771. Kamath, R.S., and Ahringer, J. (2003). Genome-wide RNAi screening in Caenorhabditis elegans. Methods (San Diego, Calif) 30, 313-321. Koca, Y., Housden, B.E., Gault, W.J., Bray, S.J., and Mlodzik, M. (2019). Notch signaling coordinates ommatidial rotation in the Drosophila eye via transcriptional regulation of the EGF- Receptor ligand Argos. Scientific reports 9, 18628. Kornfeld, K., Guan, K.L., and Horvitz, H.R. (1995). The Caenorhabditis elegans gene mek-2 is required for vulval induction and encodes a protein similar to the protein kinase MEK. Genes & development 9, 756-768. Kumar, J.P. (2012). Building an ommatidium one cell at a time. Developmental dynamics : an official publication of the American Association of Anatomists 241, 136-149. Kumsta, C., and Hansen, M. (2012). C. elegans rrf-1 mutations maintain RNAi efficiency in the soma in addition to the germline. PloS one 7, e35428. Lackner, M.R., Kornfeld, K., Miller, L.M., Horvitz, H.R., and Kim, S.K. (1994). A MAP kinase homolog, mpk-1, is involved in ras-mediated induction of vulval cell fates in Caenorhabditis elegans. Genes & development 8, 160-173. Miller, L.M., Gallegos, M.E., Morisseau, B.A., and Kim, S.K. (1993). lin-31, a Caenorhabditis elegans HNF-3/fork head transcription factor homolog, specifies three alternative cell fates in vulval development. Genes & development 7, 933-947. Myers, T.R., and Greenwald, I. (2007). Wnt signal from multiple tissues and lin-3/EGF signal from the gonad maintain vulval precursor cell competence in Caenorhabditis elegans. Proceedings of the National Academy of Sciences of the United States of America 104, 20368- 20373. Schmitz, C., Kinge, P., and Hutter, H. (2007). Axon guidance genes identified in a large-scale RNAi screen using the RNAi-hypersensitive Caenorhabditis elegans strain nre-1(hd20) lin- 15b(hd126). Proceedings of the National Academy of Sciences of the United States of America 104, 834-839. Shaye, D.D., and Greenwald, I. (2002). Endocytosis-mediated downregulation of LIN-12/Notch upon Ras activation in Caenorhabditis elegans. Nature 420, 686-690. Shaye, D.D., and Greenwald, I. (2005). LIN-12/Notch trafficking and regulation of DSL ligand activity during vulval induction in Caenorhabditis elegans. Development (Cambridge, England) 132, 5081-5092. Sijen, T., Fleenor, J., Simmer, F., Thijssen, K.L., Parrish, S., Timmons, L., Plasterk, R.H., and Fire, A. (2001). On the role of RNA amplification in dsRNA-triggered gene silencing. Cell 107, 465-476.

Simmer, F., Moorman, C., van der Linden, A.M., Kuijk, E., van den Berghe, P.V., Kamath, R.S., Fraser, A.G., Ahringer, J., and Plasterk, R.H. (2003). Genome-wide RNAi of C. elegans using the hypersensitive rrf-3 strain reveals novel gene functions. PLoS biology 1, E12. Stinchcomb, D.T., Shaw, J.E., Carr, S.H., and Hirsh, D. (1985). Extrachromosomal DNA transformation of Caenorhabditis elegans. Molecular and cellular biology 5, 3484-3496. Sundaram, M.V. (2013). Canonical RTK-Ras-ERK signaling and related alternative pathways. WormBook : the online review of C elegans biology, 1-38. Tan, P.B., Lackner, M.R., and Kim, S.K. (1998). MAP kinase signaling specificity mediated by the LIN-1 Ets/LIN-31 WH transcription factor complex during C. elegans vulval induction. Cell 93, 569-580. Underwood, R.S., Deng, Y., and Greenwald, I. (2017). Integration of EGFR and LIN-12/Notch Signaling by LIN-1/Elk1, the Cdk8 Kinase Module, and SUR-2/Med23 in Vulval Precursor Cell Fate Patterning in Caenorhabditis elegans. Genetics 207, 1473-1488. Wang, B.B., Muller-Immergluck, M.M., Austin, J., Robinson, N.T., Chisholm, A., and Kenyon, C. (1993). A homeotic gene cluster patterns the anteroposterior body axis of C. elegans. Cell 74, 29-42. Wilkinson, H.A., and Greenwald, I. (1995). Spatial and temporal patterns of lin-12 expression during C. elegans hermaphrodite development. Genetics 141, 513-526. Yoo, A.S., and Greenwald, I. (2005). LIN-12/Notch activation leads to microRNA-mediated down-regulation of Vav in C. elegans. Science (New York, NY) 310, 1330-1333. Zhuang, J.J., and Hunter, C.P. (2011). Tissue specificity of Caenorhabditis elegans enhanced RNA interference mutants. Genetics 188, 235-237.

Appendix

Appendix A: Introduction

Vulval competence group

The VPCs are polarized epithelial cells with apical and basolateral plasma membrane domains separated by adherens junctions. The apical domain faces the ventral side of the hermaphrodite, and the basolateral domain is situated below the gonad. Six VPCs, called P3.p-

P8.p, are born during the L1 stage. These VPCs form the “vulval competence group” and can respond to different signals to adopt different vulval fates. In normal development, P5.p, P6.p and P7.p will adopt vulval fates and their descendants will become the vulva, while P3.p, P4.p and P8.p will divide once and their daughters will fuse with the major hypodermal syncytium, hyp7 (Sulston and Horvitz, 1977).

Studies showed that the Hox gene lin-39 is required cell autonomously to keep the VPCs from fusing with hyp7, thereby keeping them competent to receive patterning signals. eff-1 promotes the cellular fusion of P1.p, P2.p and P9-P11.p with hypodermis (Clark et al., 1993;

Wang et al., 1993). LIN-39 represses the transcription of fusogen gene eff-1 in the VPCs to prevent them from fusing with the hypodermis (Shemer and Podbilewicz, 2002). lin-39 expression is regulated by WNT and EGFR signaling in L2 stage. cwn-1 and egl-20, which are

WNT ligands, are expressed by the surrounding tissues and prevent premature VPC fusion with the hypodermis (Gleason et al., 2006; Myers and Greenwald, 2007).

VPC specification

In the L3 stage, the VPCs adopts one of three fates, primary (1o), secondary (2o), or tertiary (3o). In normal development, the six VPCs invariantly adopts a 3o- 3o-2o-1o-2o-3o pattern

(Sulston and White, 1980; Sternberg and Horvitz, 1986). The tertiary cells will divide once and

fuse with the hypodermis, while the primary and secondary cells will divide more than once and generate the vulva.

The inductive signal, LIN-3/EGF, is sent by the anchor cell in the somatic gonad beginning in the L2 stage, and activates the LET-23/EGFR pathway. VPC fates are determined in the L3 stage. P6.p adopts the primary fate and expresses a lateral signal composed of LAG-2,

APX-1, and DSL-1 (Chen and Greenwald, 2004). The lateral signal activates LIN-12/Notch in the flanking VPCs, P5.p and P7.p, causing them to adopt a secondary fate (Figure 1). Therefore, the appropriate adoption of primary and secondary cell fate is essential in normal vulval development. Disruption of these pathways and the stereotypical 3o-3o-2o-1o-2o-3o pattern often lead to a visible mutant vulval phenotype, such as Vulvaless (Vul) in the case where there is not inductive signal, or Multivulva (Muv) in the case where there is ectopic induction.

EGFR-Ras-ERK signaling in P6.p and its regulators

The subcellular localization of EGFR on P6.p is crucial for its ability to respond to lin-

3/EGF from the gonad. In normal development, EGFR is mostly localized on the basolateral membrane which is in close proximity to the ligand source. Mislocalizing EGFR to the apical domain by either mutating LET-23 itself or removing the ERM (Ezrin, Radixin, Moesin) complex (lin-2/lin-7/lin-10) can result in a Vul phenotype (Kaech et al., 1998).

Upon ligand binding, EGFR autophosphorylates and recruits the scaffolding protein

SEM-5/GRB2 and the guanine nucleotide exchange factor LET-341/SOS-1 (Sundaram, 2013;

Clark et al., 1992; Chang et al., 2000). SOS-1 binds and activates LET-60/Ras, which initiates a phosphorylation cascade of LIN-45/RAF, MEK-2/MEK, and MPK-1/ERK (Han and Sternberg,

1990; Han et al., 1993; Kornfeld et al., 1995; Lackner et al., 1994). Activation of the EGFR-Ras-

ERK pathway result in activation of many downstream effectors that promote vulval fates such

as the forkhead-like transcription factor LIN-31 and ETS domain containing protein LIN-1. LIN-

1 plays a critical role in establishing vulval cell fate (Ferguson and Horvitz, 1985; Miller et al.,

1993; Beitel et al., 1995). It is required for both the inhibition and activation of primary cell fate

(Beitel et al., 1995; Tan et al., 1998; Jacobs et al., 1998). When MAPK is activated in P6.p,

MPK-1/ERK phosphorylates LIN-1 to relieve the LIN-1 mediated transcriptional repression.

LIN-31 was found to form a complex with LIN-1 and is required to promote as well as suppress vulval cell fate.

There are cell non-autonomous negative regulators of the pathway, mainly the synMuv genes (class A, B and C). Through mosaic analysis and tissue-specific rescue transgene experiments, they were found repress ectopic ligand production in the hypodermis (Herman and

Hedgecock, 1990; Myers and Greenwald, 2005). There are also likely cell-autonomous negative regulators. Through suppressor screens, studies identified lst-4, sli-1, ark-2, gal-1, lip-1, gyp-23 unc-101, and apm-1 (Sundaram, 2013). Some of these factors are important in ensuring the invariant 3o- 3o-2o-1o-2o-3o pattern of cell fates, which will be detailed below.

LIN-12/Notch signaling and its regulators

(Please refer to Chapter 1 for introduction of the signaling cascade and its components.)

After P6.p adopts a primary cell fate, it expresses the lateral signal encoded by DSL ligands, namely two transmembrane ligands LAG-2 and APX-1, and a secreted ligand DSL-1.

These ligands activate LIN-12 in P5.p and P7.p, and as a result they adopt a secondary cell fate.

Several LIN-12 targets play a role in vulval development, with a lot of them being negative regulators of EGFR-Ras-ERK pathway (Yoo et al., 2004; Berset et al., 2001). One LIN-12/Notch target is a MAP kinase phosphatase. Such proteins are protein threonine/tyrosine phosphatases

that inactivate MAP kinases by dephosphorylation. lip-1 encodes a MAP kinase phosphatase and is expressed in a LIN-12 dependent manner (Berset et al., 2001).

Another class of LIN-12/Notch targets includes microRNAs. MicroRNAs are small non- coding regulatory RNAs that mediate post-transcriptional down-regulation of genes. mir-61 is a microRNA that is a direct transcriptional target of LIN-12. It targets and downregulates vav-1, the ortholog of vav oncogene, a negative regulator of lin-12 activity in P5.p and P7.p (Yoo and

Greenwald, 2005).

LIN-12/Notch is also negatively regulated for homeostasis. SEL-10/Fbw7 is one well- characterized example. It induces the degradation of LIN-12 intracellular domain via the canonical Cdc4 PhosphoDegron in the region containing a PEST sequence (Hubbard et al., 1997; de la Cova and Greenwald, 2012; Deng and Greenwald, 2016). This gene came from a screen for suppressors of a lin-12 hypomorphic allele (Sundaram and Greenwald, 1993a and b). Additional negative regulators of LIN-12/Notch were identified through screening for kinase genes that enhance lin-12 hypermorphs (Deng et al., 2019). Components of a modulator of the Mediator complex (cdk-8, sur-2/med23, and lin-25/med24) are required for lateral signal expression and to downregulate LIN-12/Notch in P6.p (Shaye and Greenwald, 2002; Underwood et al., 2017).

LIN-12/Notch can also be negatively regulated to downregulate LIN-12/Notch activity. One example is the downregulation of LIN-12 mediated by EGFR activation in P6.p, which will be detailed below.

Interplay of Notch and EGFR signaling in C. elegans VPC specification: The antagonism of

LIN-12/Notch pathway by LET-23/EGFR activation

As mentioned above, normal hermaphrodites have an invariant vulval fate pattern of 3o-

3o-2o-1o-2o-3o. This is a result of integrating EGF and DSL signaling inputs. Here I will focus mainly on the crosstalk between EGFR and Notch pathway.

C. elegans VPCs receive a graded LIN-3 signal from the anchor cell, with P6.p receiving the highest level of lin-3, and P5.p and P7.p receiving a low level of lin-3. However, P5.p and

P7.p always adopts a secondary cell fate, so this raises the question as to how the stereotypical

2o-1o-2o cell fate is achieved. There are studies indicating that this is achieved by 1) promoting

LIN-12 activity in P5.p and P7.p, 2) downregulating EGFR activity in P5.p and P7.p, and 3) downregulating LIN-12 activity in P6.p. (Please refer to chapter 1 for how LIN-12 activity is promoted and how EGFR activity is downregulated in P5.p and P7.p.)

lin-12 is transcribed in all the VPCs before induction (Wilkinson and Greenwald, 1995).

Based on the expression pattern of a lin-12 transcriptional reporter, it was concluded that LIN-12 is not transcriptionally downregulated (Wilkinson and Greenwald, 1995). However, a rescuing

LIN-12::GFP transgene revealed that protein was absent in P6.p, suggesting post-transcriptional downregulation (Levitan and Greenwald, 1998). Such downregulation is dependent on EGFR activation, since mutating components of the EGFR pathway abolishes the downregulation

(Shaye and Greenwald, 2002).

Following from these observations, removal of a 15 amino acid segment of the LIN-12 intracellular domain called the Downregulation Targeting Sequence (DTS) suggested that it allowed LIN-12 to persist in P6.p and can inhibit lateral signaling; in addition, expression of

LIN-12 ectodomain in P6.p is sufficient for such inhibition (Shaye and Greenwald, 2005) (Figure

2).

Additional observations suggested that LIN-12 is post-translationally downregulated in

P6.p. After the DTS was identified, a more detailed look into that region reveals a stretch of 15 amino acids that contains a clathrin-mediated endocytosis targeting signal sequence, a di-leucine motif. Mutation of the di-leucine resulted in the stabilization of LIN-12 on the membrane in

P6.p. Mutations in serines, threonines, and lysines, which are residues that could be modified to prime LIN-12 for endocytosis, result in abnormal accumulation of LIN-12::GFP in punta or mislocalization of LIN-12::GFP to basolateral membrane (Shaye and Greenwald, 2002) (figure

3a). Additional screens for trans factors that act on DTS to downregulate LIN-12 revealed the role of endocytic proteins and ubiquitin ligases such as alx-1 and wwp-1, which are C. elegans ortholog of yeast Bro1p and Su(dx) respectively (Shaye and Greenwald, 2005). ALX-1 interacts with ESCRT protein complex and promotes MVE formation, which is important for protein recycling or degradation. wwp-1 is a Nedd4 family of E3 ubiquitin ligases, its ortholog

Suppressor of Deltex Su(dx) was a known negative regulator of Notch. However, these factors do not seem to affect the LIN-12 internalization step, since knocking down these factors does not affect lateral signaling. All these experiments which looked at expression levels were done using multicopy simple arrays such that expression was often variable and higher than endogenous levels. Moreover, all visualizations of LIN-12::GFP were done using fixation and antibody staining.

Conservation of Notch endocytic regulation mediated by EGFR

The di-leucine based motif defined in the LIN-12 DTS is highly conserved in all known nematode LIN-12/Notch proteins. A DTS-like sequence is not found in Drosophila or

mammalian Notch proteins, but vertebrate Notch1 proteins do contain a highly conserved di- leucine-like motif (Zheng et al., 2013). In addition, it is possible that the intracellular domain of

Notch proteins contains other potential endocytic signals. Given that most endocytic signals consist of short degenerate motifs of four to seven residues, only two or three are critical for function (Bonifacino and Traub, 2003).

Interplay between Notch and Ras in other organisms

There are a few examples in mammalian cells and Drosophila of the interplay between

Ras and Notch. In mammalian adult brain, niches that maintain a source of neural stem cells and neural progenitor cells are located in the subventricular zone (SVZ). Notch regulates neural stem cell identity while EGFR regulates neural progenitor cell proliferation and migration (Hitoshi et al., 2002). The interplay of the two pathways maintain the balance between neural stem cell and neural progenitor cell numbers by EGFR promoting Notch downregulation via Numb (Aguirre et al., 2010). Ras and Notch interplay is also important in disease context, especially in cancer, with studies demonstrating a cooperative or antagonistic relationship between the two in different contexts (Guo et al., 2010).

Another well-known example of Notch and Ras interplay is in Drosophila eye ommatidia development. A mature ommatidium consists of 8 photoreceptors (R1 to R8), 4 cones cells, 7 pigments cells and a mechanosensory complex (Kumar, 2012). These cells are specified by a combination of Notch and EGFR signaling. Initially, Notch lateral inhibition specifies a single cell to become R8, and activates the transcription of EGF. EGF then induces the immediate neighbors of R8 to differentiate into R2 and R5 photoreceptor pair. These receptors then express

EGF and induces their immediate neighbor to become R3 and R4. R1 and R6 are specified in a similar manner. These cells then activate Notch as well as RTKs (EGFR and Sevenless) to

specify R7. In every step, there is cross-talk between Notch and EGFR pathways. For example,

Notch regulates aos transcription, an inhibitory ligand to EGFR, in R4 precursor during R3/R4 specification where high Notch activity specifies R4 fate (Fanto and Mlodzik, 1999; Koca et al.,

2019).

Original aims for this study

1) A detailed cis-analysis of LIN-12 DTS region

Previous cis analysis was done by expressing transgenes in a high copy number and

visualized by staining. Such “simple” arrays can vary properties over time and staining is a

tedious process. Moreover, due to limitations in transgenesis and cloning, the serines and

lysines were mutated as a group. I will examine single copy insertion transgenes at a specific

chromosomal site, which will allow me to compare LIN-12 localization among mutants with

single amino acid changes in the DTS with live imaging. This will also allow me to identify

residues that are required for either internalization or degradation of LIN-12.

2) A trans factor screen for regulators in LIN-12 downregulation in P6.p and its descendants

a) Lack of LIN-12 downregulation confers to change in cell fate, which can be detected

with cell fate markers or on a gross level. I will conduct an RNAi screen looking for

inappropriate cell fate adoption as well as subcellular localization of LIN-12. As positive

controls, I will test the effect of lowered levels of kinases downstream of EGFR, such as

Braf, MEK and ERK. I will then look at the expression and localization of candidate

genes obtained from the RNAi screen in response to EGFR activation.

b) In C. elegans, it remains unknown whether or how LIN-12 is trafficked via different

pathways in VPCs with and without EGFR activation or when LIN-12 is activated. In the

case of LET-23-mediated LIN-12 endocytosis, I will conduct mutational analysis of LIN-

12 to identify region that is necessary and sufficient for LIN-12 downregulation, as well

as to identify how specific residues contribute to the endocytic regulation of LIN-12. In

parallel, I will conduct RNAi on genes involved in different steps of trafficking such as

internalization, early to late endosome transition, apical and basolateral recycling etc.

This allows me to identify the endocytic regulation of LIN-12 in VPCs with and without

EGFR activation as well as that of activated LIN-12.

Introduction Figures

Figure 1. VPC specification. In L1, 6 VPCs (P3.p to P8.p) are born. In L2, specified AC secretes LIN-3/EGF and activates the EGFR pathway in P6.p, which then adopts 1o fate. P6.p then expresses the Notch ligands LAG-2, APX-1 and DSL-1 to activate LIN-12/Notch signaling in the neighboring cells, P5.p and P7.p, which adopt the 2o fate. The VPCs that do not receive any signal will adopt the 3o fate. VPCs adopting 1o and 2o vulval fates will generate descendants that form a functional vulva. VPCs adopting the 3o fate will divide once and produce daughters that fuse with the hypodermis.

Figure 2. LIN-12 expression in VPCs and their descendants. LIN-12 is expressed in all VPCs initially. Its expression starts to go away in L2 molt in P6.p, when it starts to adopt 1o fate. In the Pn.px stage, expression is further downregulated in P6.px, and goes away in the tertiary VPCs as they fuse with the hypodermis. On the right are live images of P5.p, P6.p, P7.p and their descendants. Green is LIN-12::GFP from wgIs72, an integrated lin-12g::GFP fosmid. Nuclear red is 1o cell fate marker arIs222[lag-2p::2xnls-tagRFP].

Figure 3. LIN-12 internalization mediated by the Downregulation Targeting Sequence (DTS) (Photomicrographs are from Shaye and Greenwald, 2002). LIN-12 RAM domain. DTS is highlighted in red. Serine, threonine and lysine residues mutated in Shaye and Greenwald (2002) are underlined (Top panel). LIN-12::GFP is downregulated inP6.px and localized in puncta, while LIN-12(DTS)::GFP is stabilized on P6.px membrane, and seems to fail to be internalized (Bottom panel). These images are from stained animals.

100

Figure 4. LIN-12 downregulation mediated by alx-1 and wwp-1 (Figure from Shaye and Greenwald, 2005). These photomicrographs are ventral views of VPCs. Both alx-1 and wwp-1 knockdowns cause accumulation of LIN-12::GFP in puncta (left panels). However, it did not result in a lateral signaling defect as AJM-1 is still visible in P5.pxx and P7.pxx, meaning both adopted a vulval cell fate and their descendants did not fuse with Hyp7 (right, top 3 panels). This indicates that neither alx-1 and wwp-1 is not required for LIN-12 internalization but perhaps are involved in routing LIN-12 to endosomal compartments. These images are from stained animals.

101

Appendix B: Results

102

Issues with the trans-factor RNAi screen

To identify trans-factors that are involved in EGFR-mediated LIN-12 downregulation, we planned to perform a targeted screen of conserved kinases and E3 ubiquitin ligases using the

Ahringer library, some of which are essential genes (Kamath and Ahringer, 2003). We also planned to perform a targeted screen for endocytic factors involved in general LIN-12 regulation, a lot of which are also essential genes. For that reason, we first generated a VPC-specific RNAi strain to bypass lethality. We created a single copy insertion of lin-31p::rde-1 to rescue rde-

1(ne300), a null allele, in the VPCs. RDE-1 is an Argonaut protein required for RNAi and rde-

1(ne300) is a null allele of rde-1 that results in animals that are refractory to RNAi (Tabara et al.,

1999). We used single-copy insertion to eliminate variability between animals in an RNAi experiment, which happens when extrachromosomal arrays were used due to their variable expression between animals. To test for the tissue specificity of the RNAi strain, we performed

RNAi against genes with different tissue focus, for example unc-22 for muscle and pos-1 for germline, and found these tissues are refractory to RNAi (Figure 1a).

To assess if the VPCs are sensitive to RNAi, we performed several positive controls for the screen. There are three cellular foci for the screens. One is P6.p, where we will be screening for factors in EGFR-mediated LIN-12 downregulation; another one is P5.p and P7.p, where we will be screening for endocytic factors regulating LIN-12 in a Notch signaling active context; and the third is the prospective 3o VPCs, where we will be screening for factors regulating LIN-

12 when LIN-12 is not active. We first performed feeding RNAi against lin-1, an ortholog of the

ELK-1 transcription factor that is expressed in all VPCs and plays a role in activating and repressing vulval cell fate (Beitel et al., 1995; Tan et al., 1998; Jacobs et al., 1999). The lin-1 null mutant has a highly penetrant Multivulva phenotype. We found that VPCs in rde-1(0);

103

arTi106[lin-31p::rde-1] nre-1 lin-15b are indeed sensitive to RNAi based on lin-1(RNAi) causing a Multivulva phenotype, suggesting that RNAi capability was restored to the prospective

3o VPCs (Figure 1a). We also performed lin-12 RNAi and observed a highly penetrant vulval defect (Pvl) as well as loss of expression of a lin-12 transcriptional target, ar116[lst-

5p::2xnlsYFP] suggesting that P5.p and P7.p are sensitive to RNAi. We next assessed whether

P6.p is sensitive to RNAi at a level enough for our screen. We performed RNAi against components of the EGFR pathway, such as let-23, let-60, lin-45, mpk-1, hoping to use them as positive controls for the screen (Figure 1b). Unfortunately, we were not able to detect a high penetrance of the expected Vulvaless phenotype, which can potentially create a lot of false negatives in the screen (Figure 1c). This could be due to either issues with RNAi clones themselves, or the insufficient level of rde-1 rescue.

To distinguish between these scenarios, we compared two strains, one is a VPC-specific

RNAi strain, and another one is a pan RNAi sensitive strain, with the Ras pathway RNAi clones listed above. We performed the same experiment with the two strains in parallel, and found no improvement on the effectiveness of EGFR pathway RNAi (Figure 1e). We also performed

RNAi on different strains with different RNAi sensitizers such as eri-1 and rrf-3, but found no improvement on effectiveness of EGFR pathway RNAi (Figure 1d) (Simmer et. al., 2003;

Zhuang and Hunter, 2011). These results suggest that P6.p is more refractory to RNAi. To test the hypothesis that rde-1 rescue might be insufficient, we also looked at expression of lin-31 promoter in several single copy transgenes (some described in section below), and observed decreased lin-31p expression in P5.p, P6.p and P7.p in independent single copy transgenes. We concluded that P6.p in this strain is likely not sensitive enough for screening purpose and that

EGFR pathway components are somehow refractory to RNAi in P6.p. While there are ways to

104

improve the strain which I will discuss in later sections, we decided to put the screen on hold to focus on the cis-analysis of LIN-12, where we encountered more pressing issues as discussed below.

Weak and inconsistent expression from single copy lin-31p::lin-12 transgenes

The original LIN-12 cis-analysis were done by assessing transgenes that were high-copy number simple arrays, where expression is often variable and higher than endogenous level (Fire and Waterston, 1989; Stinchcomb et al., 1985). Expression was assayed by fixing the animals and visualized by GFP antibody staining. Moreover, most of the transgenes used lin-12 genomic sequence, which allows for expression in the gonad and might obscure visualization of basolateral domain on P5.p, P6p and P7.p. Therefore, we decided to generate transgenes using lin-31p, a VPC specific promoter, in a single copy manner to allow appropriate comparison between transgenes. Here, we used lin-12 cDNA, with GFP tagged C-terminally. Previous transgenes used lin-12 genomic sequence, with GFP replacing a 17aa XhoI fragment, which is located upstream of the PEST domain and downstream of ANK repeats. The reason we used the

C-terminal tag here is because based on previous data, Region 1 is sufficient for LIN-12 downregulation. More importantly, deleting the DTS or mutating the di-leucine in Region 1 only can stabilize LIN-12 in P6.p and descendants, suggesting that the motifs mediating downregulation lies within Region 1 (Figure 4a) (Shaye and Greenwald, 2002; Shaye and

Greenwald, 2005). The implication of this assumption will be discussed later.

We generated single copy insertions of lin-31p::lin-12-GFP::unc-54 3’UTR. We observed expression at the L2 stage (based on qualitative assessment of gonad extension and

VPC size), however it disappears in the L2 molt, when EGFR-mediated LIN-12 downregulation occurs. We next tried expressing Region 1 of lin-12, which is necessary and sufficient for EGFR-

105

mediated LIN-12 downregulation in P6.p. We hoped that without other negative regulatory sequences on LIN-12, such as the PEST domain and the “Y region”, LIN-12 might be stabilized and therefore enough for visualization (Hubbard et al., 1997; Deng and Greenwald, 2016).

However, I could not observe any expression from single copy transgene of wildtype version of

Region 1, as well as Region 1 with the di-leucine motif mutated.

Over the course of the study, we also tried different fluorescent proteins, such as LIN-

12::mNeonGreen and Region 1::mClover, given their reported brightness in vivo (Hepper et al.,

2016; Hosteller et al., 2017). We observed perinuclear accumulation of LIN-12::mNeonGreen and no expression from Region 1::mClover. We re-injected some of these constructs to make multicopy complex arrays, and we were able to see weak and unexpected expression pattern of

Region 1::mClover and lin-12::GFP (expression pattern of lin-31p will be discussed below).

The expression data is summarized in figure 2a. We concluded that single copy lin-31p is likely not strong enough to drive visible expression of LIN-12::GFP at the stage we want.

To troubleshoot this problem with expression level, we first tried a different promoter, egl-17p (Burdine et al., 1998) to express lin-12 and lin-12(ΔDTS). Based on observation from previous studies, the promoter is on in P6.p and its descendants and was used in Shaye &

Greenwald (2002 and 2005). Unfortunately, we did not observe expression using this transgene either (Figure 2a). Next, we decide to combine two single copy insertions of lin-31p::lin-12-

GFP. We observed apical GFP expression, with GFP visible on all the VPCs at early Pn.p stage and downregulation in P6.p. However, at later Pn.p stage, it is only visible on P3.p, P4.p and

P8.p (Figure 2b). We next looked at single copy insertion lin-31p::lin-12(ΔDTS)::GFP. We observed similar expression pattern as LIN-12::GFP, where GFP is expressed in the VPCs at

106

early Pn.p stage, with lower expression in P6.p while at later Pn.p stage, expression is only visible on P3.p, P4.p and P8.p (Figure 2b).

In an attempt to further stabilize LIN-12(ΔDTS)::GFP in later stages, we assayed the transgene in a sel-10(0) background. sel-10 negatively regulates LIN-12 through the PEST domain at C-terminus of LIN-12(Hubbard et al., 1997). We observed nuclear stabilization of

GFP, but not on the apical membrane. More importantly, we did not observe GFP in P6.p.

This unexpected expression pattern of GFP being in only P3.p, P4.p and P8.p could be due to lin-31p expression. All previous transgenes using lin-31p are either multicopy number array, or driving expression of histone markers which are highly stabilized proteins. We generated lin-31p::2xnls-GFP::unc-54 3’UTR and noticed that GFP is indeed dimmer in P5.p,

P6.p and P7.p at late Pn.p stage (Figure 2c). While there are inherent issues with the lin-31p for its expression pattern and level, the more unexpected result is that lin-12(ΔDTS)::GFP is downregulated in P6.p even in early Pn.p stage and that it behaves similarly to LIN-12::GFP, even though LIN-12(ΔDTS)::GFP seems to be more stabilized, as suggested by expression in late Pn.p stage with a single copy lin-12(ΔDTS)::GFP transgene, and not with single copy lin-

12::GFP transgene.

Revisiting lin-12(ΔDTS) high copy number transgenes

The results from single copy insertion of lin-31p::lin-12(ΔDTS)::GFP suggests that 1)

LIN-12(ΔDTS) is either downregulated in P6.p due to lin-31p expression pattern, 2) LIN-

12(ΔDTS)::GFP does not reflect endogenous dynamics, 3) LIN-12 is not regulated through DTS

(or DTS alone) in P6.p, or more likely, a combination of all of the above.

107

We revisited the LIN-12(ΔDTS) high copy number transgenes that were driven under lin-

12 genomic sequence and egl-17 promoter and re-examined them in modern conditions (results summarized in Figure 3a). Previously, these transgenes were visualized by staining which might amplify the signal. Moreover, transgenes were made in an unc-4 background for the selection purpose. We assayed the same strains using live imaging, and was able to observe penetrant GFP expression in P6.p and P6.px, as well as egg laying defects, concordant with the observation from Shaye & Greenwald (2002 and 2005) (Figure 3b).

We then used the same plasmids to create transgenes that are low copy number arrays as well as high copy number arrays (similar to the original protocol) in the pha-1(e2123ts) background instead of unc-4. We did not observe any GFP expression in P6.p for egl-17p:: lin-

12(ΔDTS)::GFP in either condition. However, we did observe some GFP expression in P6.p in lin-12g::lin-12(ΔDTS)::GFP when injected as high copy number arrays (Figure 3c).

One possibility is that LIN-12(ΔDTS)::GFP expression on P6.p is very sensitive to expression level, and that the previous results with LIN-12(ΔDTS)::GFP high copy number arrays might not reflect endogenous LIN-12 regulation. Therefore, we decided to test the requirement for DTS in the endogenous context.

Endogenous lin-12(ΔDTS)::GFP expression

To investigate whether endogenous LIN-12 is regulated through the DTS in P6.p, we made lin-12(ar638), an endogenous lin-12(ΔDTS)::GFP allele using CRISPR/Cas9. There is one difference of this allele from the multicopy transgenes used in previous studies: the GFP is C- terminally tagged instead of replacing a XhoI fragment (Figure 4a) (Levitan and Greenwald,

1995; Shaye and Greenwald, 2002). The GFP in previous transgenes had been inserted in this

108

manner for ease for cloning, and such manipulation does not affect LIN-12 function, since these transgenes rescue lin-12(0). For the C-terminal GFP tag, we had previously generated a C- terminal LIN-12::GFP endogenous tag. It has no apparent lin-12 phenotypes and therefore we concluded that the position of the tag does not affect LIN-12 function.

Endogenous LIN-12(ΔDTS)::GFP seems to be functional, as the animals are healthy and fertile. Moreover, we did not observe any vulval defects that are associated with LIN-

12(ΔDTS)::GFP described previously. Even though LIN-12(ΔDTS)::GFP is expressed at a higher level than LIN-12::GFP, the key result here is that LIN-12(ΔDTS)::GFP is downregulated in

P6.p, similar to LIN-12::GFP (Figure 4b). This again suggests that the results from previous high copy number transgene may not reflect endogenous LIN-12 regulation. Alternatively, there may be some other sequence motif within the 17 aa long XhoI fragment, or some other sequences, that are important for endogenous LIN-12 downregulation in P6.p.

Endogenous lin-12 transcriptional reporter

Since it is possible that LIN-12 might be regulated in P6.p independent of DTS, we decided to revisit the assumption that LIN-12 is regulated post-transcriptionally in P6.p, given that the evidence for post-transcriptional regulation is based on high copy number transgenes which might not reflect endogenous regulation. We generated an endogenous lin-12 transcriptional reporter by inserting sl2::2xnls-GFP c-terminally to lin-12. We found that the endogenous transcriptional reporter is downregulated in P6.p and P6.px (Figure 5). This suggests that there is some level of transcriptional regulation of lin-12 in P6.p. We also assayed the reporter expression in a lin-1(0) background, and observed that the transcriptional downregulation in P6.p and its descendants was abolished. This result suggests the

109

transcriptional downregulation could be mediated by EGFR activation, since lin-1 is one of the downstream effectors of the pathway.

Endogenous lin-12(ΔDTSΔXhoI)::GFP expression

As mentioned before, there is one key difference between the transgenes used previously and ones in this study, aside from copy number. We generated an endogenous lin-12 allele identical to the transgenes used previously, deleting the XhoI fragment. Prior to excision of the self-excising cassette (SEC), we observed penetrant Egl defects as well as low penetrant Muv phenotype, similar to what was observed in previous studies. However, animals appear wildtype after SEC was excised (Figure 6). When we examined the strain for GFP expression, we did not see GFP stabilization in P6.p and its descendants. The discrepancy could be due to: 1) the presence of regulatory sequences within the SEC might enhances expression of lin-

12(ΔDTSΔXhoI)::GFP to the point that it causes a lateral signaling defect, 2) SEC excision resulted in additional mutations that rescued the defects caused by DTS and XhoI fragment deletion. We were able to confirm the DTS and XhoI fragment deletion in the strain prior to SEC excision, but unfortunately we were not able to confirm the final strain with sequencing of the entire coding region due to time constraints.

110

Results Figures

111

Figure 1. VPC-specific RNAi screen. A) Testing tissue specificity of RNAi strain. The table summarize the results from testing the tissue specificity of the rde-1 rescue strain. rde-1(0);arTi[lin-31p::rde-1] nre-1 lin- 15b is sensitive to RNA in the VPCs and refractory to RNAi in other tissues such as the somatic gonad, germline, and muscle. B) Components of LET-23/EGFR pathway (Adapted from Sundaram, 2006). LIN- 3/EGF from the anchor cell binds to LET-23/EGFR to activate downstream ras-raf-Mek- Erk cascade. Activated MPK-1/Erk phosphorylates LIN-1/ELK1 (in red) which represses vulval cell fate and other effectors such as SUR-2, LIN-25, EOR-1 and EOR-2 in green promote vulval cell fate. Yellow circle indicates genes targeted by RNAi in this study.

112

c) RNAi against EGFR pathway components in the VPC-specific RNAi strain {give the strain name and full genotype}. lin-1 RNAi affects all 6 VPCs and lin-12 RNAi affects P5.p and P7.p. Both lin-1 and lin-12 RNAi are effective and result in high penetrance of vulval defects including Muv and Pvul animals. let-23, let-60 lin-45, mpk-1 and lin-1 are components of the EGFR cascade. gsk-3 and cye-1 are regulators of LIN-45 turnover in VPCs. RNAi against EGFR pathway components and regulators results in low penetrance of hypomorphic phenotypes. None of the RNAi treatments produced a Vul phenotype that is significant (p<0.01) compared to lacZ.

113

D) RNAi on different RNAi sensitizer strains. rrf-3(pk1426) (NL2099) and eri-1(mg366) (GS4164) do not improve RNAi effectiveness in the VPCs compared to arIs131;nre-1(hd20) lin- 15b(hd126) (GS5891), referred to here as nl. lin-39 and unc-62 are components of Hox genes that controls cell cycle progression in the VPCs. lin-39 and unc-62 mutants lose VPC competence and fail to adopt vulval cell fate, resulting in a Vul phenotype.

114

E) RNAi on VPC-specific RNAi strain and pan-RNAi strain. rde-1(0); arTi106 nre-1(hd20) lin-15b(hd126)(GS8659) refers to the VPC-specific RNAi strain which contains nre-1 lin-15b in the background. arIs131 is an integrated simple array of lag-2p::2xnls-YFP (Li and Greenwald, 2010). There is no systematic difference in RNAi effectiveness between the two strains.

115

Figure 2. Single copy transgenes of lin-12 and lin-12 mutants. Table A) Summary of expression data of different single copy transgenes. Early Pn.p stage and late Pn.p stage are determined by time post egg lay, gonad arm extension and VPC size. All transgenes are inserted either by miniMos or by CRISPR-Cas9 at the LGI site (Frokjaer-Jensen et al,. 2014; Dickinson et al., 2015). LIN-12::GFP is visible in the early Pn.p stage as a single- copy transgene and only visible in late Pn.p when two lin-31p::lin-12::GFP transgenes are combined. LIN-12(ΔDTS)::GFP seems to be more stabilized than LIN-12::GFP as it is visible in the late Pn.p stage, but it is downregulated in P5.p, P6.p and P7.p, similar to 2x LIN-12::GFP.

116

B) LIN-12::GFP and LIN-12(ΔDTS)::GFP in early and late Pn.p stage. White solid line indicates P6.p. White dotted line indicates gonad. LIN-31p::LIN-12::GFP is visible in all the VPCs in early Pn.p stage, but started to be downregulated in P6.p in both LIN-12::GFP and LIN- 12(ΔDTS)::GFP. In late Pn.p stage, LIN-12::GFP is no longer visible in the VPCs. At this stage, LIN-12(ΔDTS)::GFP is visible in P4.p and P8.p, indicated by orange solid line, but downregulated in P5.p, P6.p and P7.p. All pictures were taken at 650 ms exposure time.

117

C) Single copy transcriptional reporter of lin-31p. White line indicates P6.p (top panel) and P6.px (bottom panel), orange lines indicate P4.p and P8.p (top panel) and P4.px and P8.px (bottom panel). 2xNLS-GFP downregulation in P5.p, P6.p and P7.p is visible at Pn.p stage, and becomes more apparent in Pn.px stage. Photomicrographs are orthogonal projections of Z stacks taken on the spinning disk confocal.

118

Figure 3A) Summary of lin-12(DTS)::GFP multicopy transgenes expression. Multicopy transgenes from Shaye and Greenwald (2002 and 2005) were re-examined using live imaging, including lin-12(DTS)::GFP driven by lin-12g and egl-17p. These transgenes were generated in an unc-4 (e120) background for selection and maintaining the transgenes over generations. Integrated simple arrays were generated by bombardment. For details, refer to Shaye and Greenwald (2002 and 2005). Transgenes in this study were generated in either a wildtype N2 background, or a pha-1(e2123) background. pha-1(e2123) is a temperature sensitive mutant, with 100% lethality at 25 degrees. It was used here for selection purpose. *3/5 lines showed GFP stabilization in P6.p.

119

B) Re-examination of lin-12(DTS)::GFP transgenes from Shaye and Greenwald (2002 and 2005). Strains were re-examined using live imaging. Both lines of lin-12g::lin-12(ΔDTS)::GFP integrated and extrachromosomal simple array (arEx299, arEx300, arIs77 and arIs78) shows stabilization of GFP in P6.p apical membrane, with penetrance higher observed in integrated simple array. egl-17p is a P6.p specific promoter. However, egl-17p::lin-12(ΔDTS)::GFP (arEx389 and arEx390) did not show any stabilization of GFP in P6.p.

120

C) Stabilization of LIN-12(DTS)::GFP in P6.p in transgenes generated in this study. LIN- 12(DTS)::GFP stabilization in P6.p was not observed for any of the transgenes generated with egl-17p. 3/5 lines of lin-12g::lin-12(DTS)::GFP showed GFP stabilization in P6.p. *lin-12g complex array did now show any expression in the VPCs. The number of worms examined is shown in parentheses.

121

Figure 4. Endogenous alleles of lin-12::GFP, lin-12 (ΔDTS)::GFP, lin-12(ΔDTSΔXhoI)::GFP and lin-12::sl2::2xNLS-GFP A) Schematic of LIN-12::GFP, LIN-12(ΔDTS)::GFP, LIN-12(ΔDTSΔXhoI)::GFP and LIN- 12::sl2::2xNLS-GFP. DTS and XhoI fragment are labelled in black. Region 1, 2 and 3 are as indicated in Shaye and Greenwald, 2005. Both LIN-12::GFP and LIN-12(ΔDTS)::GFP have a flexible linker and a C-terminal GFP tag. LIN-12::sl2::2xNLS-GFP is an endogenous transcriptional reporter by inserting a sl2 transplicing site between lin-12 and 2xnlsGFP and creating bi-cistronic transcript that is regulated by the lin-12 genomic sequence.

122

B) Endogenous expression of lin-12::GFP and lin-12 (ΔDTS)::GFP. Both LIN-12::GFP and LIN-12(ΔDTS)::GFP are downregulated in P6.p and its descendants. Unprocessed images were from a slice of z-stacks taken on spinning disk confocal.

123

Figure 5. Endogenous lin-12 transcriptional reporter ar630[lin-12::sl2::2xnls-GFP]. White line indicates P6.p (top panels) or P6.px (bottom three panels). 2xnls::GFP is visible in all VPCs but is downregulated in P6.p in WT animals. At Pn.px stage, GFP is further downregulated or completely downregulated. This suggests that there is transcriptional downregulation of lin-12 in P6.p. In lin-1(0) animals, GFP is no longer downregulated as expression is uniform across VPCs in both Pn.p and Pn.px stage, suggesting that the transcriptional regulation of lin-12 could be EGFR-mediated. Images are orthogonal projections from z stacks taken on spinning disk confocal

124

Figure 6. Vulval phenotypes of endogenous lin-12 (ΔDTS ΔXhoI)::GFP before and after SEC excision, and expression of lin-12 (ΔDTS ΔXhoI)::GFP after SEC excision. A) Vulval phenotypes of lin-12 (ΔDTS ΔXhoI)::GFP before and after SEC excision. Penetrant vulval phenotypes, including Egl defects, were observed in the strain prior to SEC excision. B) Expression pattern of lin-12 (ΔDTS ΔXhoI)::GFP after SEC excision. GFP is not stabilized in P6.p (top panel) and P6.px (bottom panel). White arrows indicates P5.p and P7.p nuclei (top panel) and P5.px and P7.px nuclei (bottom panel). Images were adjusted for brightness and contrast.

125

Appendix C: Discussion

126

Our goal was to investigate the cis-regulatory sequence as well as to identify trans-factors that are required for EGFR-mediated downregulation of LIN-12 in P6.p. To identify trans- factors, we generated a VPC-specific RNAi strain which allows us to screen conserved kinases and ubiquitin ligases that are essential for development. However, we encountered issues regarding the efficiency of RNAi on EGFR pathway components and on P6.p. I will further discuss the results and possible solutions to these problems given the presence of better tools. To identify cis-regulatory sequence and residues required for downregulation, we generated several single copy and multicopy transgenes of LIN-12(DTS)::GFP. However, we did not observe stabilization of GFP in P6.p as observed in Shaye and Greenwald, 2002, which we attributed to the nature of high copy number transgenes and the inherent expression pattern of the promoter we used. We later discovered another stretch of sequence that regulates LIN-12 downregulation in P6.p, in addition to DTS. I will further discuss the implication of these results in this section.

VPC-specific RNAi of EGFR pathway components

We generated a VPC-specific RNAi strain: rde-1(0); arTi106[lin-31p::rde-1] nre-1 lin-

15b. This strain is refractory to RNAi in tissues besides VPC and hypodermis, where lin-31 is weakly expressed. The VPCs are sensitive to RNAi since we observed expected phenotypes with lin-1 RNAi, which is required in the VPCs for proper cell fate adoption. However, as positive controls, we tested the efficacy of RNAi against EGFR pathway components in P6.p, and observed weak and low penetrant phenotypes in both tissue-specific RNAi strain as well as systematic RNAi strains. For example, we would expect let-23 knockdown to result in Vulvaless

(Vul) animals, but we observed high fraction of Multivulva (Muv) animals, which is indicative of incomplete let-23 knockdown (Han et al., 1995; Kornfeld et al., 1995; Singh and Han, 1995;

Gu et al., 1998).

127

There are several possible reasons, or a combination of them, that could explain the results: 1) RNAi clones used for positive control are not effective, 2) EGFR pathway components are refractory to RNAi in the VPCs, 3) P6.p is refractory to RNAi in general. While we sequenced to confirm all the clones used in the experiment, the position of where RNAi targets can affect RNAi efficiency (Sijen et al., 2001). We could potentially improve RNAi efficiency by remaking RNAi clones to target the 5’ coding sequence (Sijen et al., 2001).

Another way to get around inefficient RNAi would be to sensitize the VPCs to perturbation of

EGFR pathway by using an EGFR-Ras-ERK hypomorph.

The second possibility is that the EGFR pathway components are refractory to RNAi in the VPCs. This could be due to the presence of compensatory mechanisms that turns on when certain components of EGFR pathway are downregulated, achieved by either upregulation of other components of the Ras-ERK pathway or by utilizing non-canonical Ras-dependent pathways. For example, non-RTK receptors such as EGL-15/FGFR can stimulate Ras- ERK signaling in sex myoblast and hypodermis; LET-60/Ras can activate RGL-1 and RAL-1 GTPase independent of Raf-MEK-ERK in P5.p and P7.p.

The last possibility is that P6.p is somehow refractory to RNAi. This is conceivable in the tissue-specific RNAi strain if the rde-1 rescue is incomplete, which is likely the case here as I will further discuss in the next section. There is evidence for differential responses to RNAi in different tissues, for that reason we always use a RNAi sensitizer nre-1 lin-15b in the background which enhances RNAi in the VPCs (Schmitz et al., 2007). However, it seems unlikely that P6.p alone is refractory to RNAi as opposed to all VPCs. Therefore, the first two scenarios more likely explain our RNAi results.

128

lin-31p::rde-1 expression and potential solutions

A separate issue with the tissue specific strain is the level of rde-1 rescue in P5.p, P6.p and P7.p. The lin-31 promoter is often used as a VPC specific promoter, given its apparently uniform expression in VPCs based on multicopy transgenes using the promoter. However, as we learned from this study, multicopy transgene expression very likely does not reflect endogenous expression pattern/level or single copy transgenes. When we looked at a lin-31p::2xnls-

GFP::unc-54 3’UTR single copy transgene, we observed lower expression of GFP in P5.p, P6.p and P7.p, which was not seen before with other transgenes.

There are several ways to get around the problem. To increase level of rde-1 expression in P5.p, P6.p and P7.p, we could generate and combine several single copy insertions of the same transgene. However, one concern is that this might further sensitize RNAi in other tissues such as hypodermis where we know there is low level of lin-31p expression. An alternative solution is based on a tool recently developed in our lab, where we can achieve tissue specific expression in very high levels. Briefly, we would generate two single copy transgenes: one transgene is lin-

31p::Cre, the other transgene drives rde-1 expression under a ubiquitous promoter that is expressed in high levels, such as rps-27, and within rde-1 introns, there would be a lox-stop codon-lox sequence. When the transgene is by itself, it would not express rde-1. When the two transgenes are combined, Cre is expressed in the VPCs and will flox out the stop codon, allowing rde-1 expression only in the VPCs but at a high level.

Another way to improve rde-1 rescue is perhaps to use a hypomorphic rde-1 allele instead of rde-1(0). Many groups use rde-1(ne219) instead of rde-1(ne300) for generating tissue specific RNAi strains (Espalt et al., 2005; Qadota et al., 2007; Firnhaber and Hammarlund,

2013). rde-1(ne219) contains a single glutamate to lysine substitution while rde-1(ne300)

129

contains an early stop. rde-1(ne219) is only partially resistant to RNAi therefore might allow for less tissue specific but more effective knockdown (Watts et al., 2018; Zou et al., 2019).

DTS deletion is not sufficient for LIN-12 stabilization in P6.p and its descendants

The caveats of using high copy number transgenes has posed issues in moving the trans- factor screen forward, and more importantly, caused us to revisit the previous assumptions made in the study using these types of transgene. Moreover, we found that deletion of the DTS in a background where lin-12 is tagged C-terminally does not cause stabilization of LIN-12 or any vulval defects. However, when GFP is inserted 5’ to the PEST domain, replacing a 17aa- encoding XhoI fragment, in combination with DTS deletion, there is observed penetrant Egl defect prior to SEC excision, similar to what was observed in previous studies. If GFP is stabilized and Egl defects are still present in the final strain after SEC excision, this would suggest DTS deletion alone is not sufficient for LIN-12 stabilization and there is some additional sequence motif within the XhoI fragment that mediates downregulation. Alternatively, the insertion of GFP at XhoI site somewhat stabilizes the protein, or the insertion GFP at C-terminus destabilizes the protein so that it is more easily downregulated. The way to truly test the requirement of the XhoI fragment is to 1) delete the fragment along with DTS in a background where lin-12 is C-terminally tagged, 2) insert GFP at XhoI site, but keep the XhoI fragment instead of replacing it (Deng and Greenwald, 2016). However, when we looked at the final strain, the Egl defect is no longer present. The potential causes for this are listed in the next section.

130

A Revised Model of LIN-12 downregulation in P6.p and future directions

In this study, we also generated a endogenous transcriptional reporter ar641[lin-

12::sl2::2xnls-GFP]. We showed that transcription is downregulated in P6.p and further downregulated in its descendants. Transcriptional downregulation of lin-12 is potentially mediated by EGFR pathway, since in a lin-1(0) background, downregulation is abolished. To confirm whether lin-12 transcriptional downregulation is really mediated by EGFR activation, one could assay the endogenous transcriptional reporter in various let-23/EGFR pathway mutants.

Moreover, we found that endogenous LIN-12(ΔDTS)::GFP is downregulated in P6.p and its descendants. However, we did notice a general stabilization of endogenous LIN-

12(ΔDTS)::GFP, meaning that DTS does mediate protein stability at some level. When we delete both the DTS and the XhoI fragment endogenously, we observed penetrant Egl phenotype and low penetrance of Muv phenotype in the strain, but only prior to SEC excision. The reason for why these phenotypes are not present in the final strain after SEC excision could be: 1) The presence of regulatory sequences within the SEC might enhances expression of lin-

12(ΔDTSΔXhoI)::GFP to the point that it causes a lateral signaling defect, 3) SEC excision resulted in additional mutations that rescued the defects caused by DTS and XhoI fragment deletion.

131

Appendix D: Materials and Methods

132

C. elegans Strains

Strain names are listed in Supplemental table 1. C. elegans strain N2 and GS6014 (pha-

1(e2123ts) were used for generating transgenes by microinjection.

The following transgenes were used in this section:

arIs131[lag2p::2xNL-YFP] is an integrated simple array containing a lag-2 red reporter.

It is expressed in P6.p and its descendants as well as the AC and the distal tip cells.

arEx299 and arEx300[lin-12g::lin-12(ΔDTS)::GFP] are extrachromosomal simple arrays containing lin-12(DTS)::GFP driven under the lin-12 genomic regulatory sequence

arIs77 and arIs79 [lin-12g::lin-12(ΔDTS)::GFP] are integrated simple arrays of arEx299 and arEx300 which contains lin-12(Dts)::GFP driven under the lin-12 genomic regulatory sequence.

arEx389 and arEx390[egl-17p::lin-12(ΔDTS)::GFP] are extrachromosomal simple arrays of lin-12(ΔDTS)::GFP driven under egl-17p, a P6.p specific promoter.

Method Details

RNAi

RNAi clones were obtained from the Ahringer library using HT115 derived bacterial strains expressing C. elegans gene sequences (Kamath and Ahringer, 2003). L1 RNAi were conducted at 25oC. Briefly, eggs were prepared from hermaphrodites maintained at 20oC using a bleach/sodium hydroxide protocol and placed on plates without OP50 to allow to hatch and starve. Next day, hatched and synchronized L1s are plate onto the RNAi plates with bacterial.

RNAi experiments were conducted at 25oC. Vulval phenotypes were assessed on DIC. Animals

133

with additional pseudovulvae were scored at Multivulva (Muv). Animals with no vulval opening were scored as vulvaless (Vul). Animals with a single protruding vulval were scored as Pvul.

Transgenesis for lin-12 cis-analysis

All single copy insertion transgene of wildtype lin-12 or mutant lin-12 were generated by random insertion using miniMos or targeted insertion on the LGI site using CRISPR-cas9 (

Frokjaer-Jensen et al,. 2014 ; Dickinson, 2015). For miniMos, miniMos-based vectors were injected into N2 at 10 ng/ul, pGH8 (Prab-3::mCherry) at 10 ng/ul and pCFJ90 (Pmyo-

2:mCherry) at 2.5 ng/ul, pCFJ601 (Peft-3:mos1 transposase) at 50 ng/ul and PMA122

(Phsp16.41::peel-1) at 10 ng/ul. For LGI site CRISPR insertions, homology templates were injected into N2 at 10 ng/ul, pAP82 at 50 ng/ul, pGH8 at 10 ng/ul, pCFJ90 at 2.5 ng/ul. All plasmids used for LGI site CRISPR insertions were prepared using Invitorgen’s Purelink miniprep kit.

All extrachromosomal arrays of wildtype or mutant lin-12 were generated by microinjections into pha-1 background as simple or complex arrays (Mello et al,. 1990; Kelly et al., 1994; Granato et al., 1994). For simple arrays, plasmids were injected at 5 ng/ul, pBX at 50 ng/ul, pBlueScript at 25 ng/ul, and ttx-3p:tagRFP @ 20ng/ul. For complex arrays, linearized plasmid were injected at 2 ng/ul, N2 genomic DNA at 50 ng/ul, pBX-1 @ 1 ng/ul, and ttx-

3p::tagRFP at 2 ng/ul. Injected P0s were kept at 15oC for 4 days, then shifted to 25oC for 4 days.

Independent transgenic lines were isolated from F2s.

Assessment of LIN-12(ΔDTS)::GFP stabilization in P6.p in transgenes

134

Worms with single copy and multicopy transgenes were scored on the fluorescence microscope at 650 ms exposure time in the GFP channel. Animals were scored at early or late

Pn.p stage, which is determined by gonad extension and VPC size.

Generation of lin-12::GFP, lin-12(ΔDTS)::GFP and lin-12(ΔDTSΔXhoI)::GFP CRISPR insertion

To generate ar624[lin-12::GFP], N2 animals were injected with homologous repair template at 50 ng/ul, 2 sgRNAs at 25 ng/ul each, along with pCFJ90 at 2.5 ng/ul and pGH8 at 10 ng/ul. All plasmids were purified with Invitrogen PureLink mini-prep kit. Successful integrants were isolated and had self-excising cassette removed according to protocol described by

Dickinson et al., 2015

To generate ar638[lin-12(ΔDTS)::GFP, N2 animals were injected with repair template at

50ng/ul, 3 sgRNAs at 17ng/ul each, along with pCFJ90 at 2.5 ng/ul and pGH8 at 10 ng/ul.

To generate ar650[lin-12(ΔDTSΔXhoI)::GFP], N2 animals were injected with repair template at 50 ng/ul, 3 sgRNAs at 17ng/ul each, along with pCFJ90 at 2.5 ng/ul and pGH8 at 10 ng/ul.

Assessment of LIN-12(ΔDTS)::GFP stabilization in P6.p in CRISPR insertions

135

arSi10[lin-31p::2xnls-GFP] , ar624[lin-12::GFP], ar638 [ lin-12(ΔDTS)::GFP] were imaged on Zeiss spinning disk confocal dual camera system at 40x magnification at 650 ms exposure time in the GFP channel. Animals were at Pn.p and Pn.px stage.

ar650[lin-12(ΔDTSΔXhoI)::GFP] were scored on the fluorescence microscope at 650 ms exposure time in the GFP channel. Animals were at Pn.p and Pn.px stage.

Table 1. Strain List Strain Genotype Source GS5891 arIs131; nre-1(hd20) lin-15b(hd126) Xinyong Zhang GS4164 eri-1(mg366) Caenorhabditis Genetics Center NL2099 rrf-3(pk1426) Caenorhabditis Genetics Center GS3126 unc-4(e120) arEx299 Dan Shaye GS3127 unc-4(e120);arEx300 Dan Shaye GS3173 unc-4(e120);arIs77 Dan Shaye GS3174 unc-4(e120);arIs78 Dan Shaye GS3363 unc-4(e120);arEx389 Dan Shaye GS3364 unc-4(e120);arEx390 Dan Shaye GS8800 arTi238[lin-31p::lin-12::GFP::unc-54 This study 3’UTR] GS8801 arSi22[lin-31p::lin-12(ΔDTS)::GFP] This study GS8802 arSi23[lin-31p::lin-12::GFP]; arTi238 This study GS8711 arSi10[lin-31p::2xnls::GFP::unc-54 3’UTR] This study GS8976 ar624[lin-12::GFP] This study GS9031 ar630 [lin-12::sl2::2xnls::GFP] This study GS9142 ar638[lin-12(ΔDTS)::GFP] This study GS9527 ar650[lin-12(ΔDTSΔXhoI)::GFP] This study GS9031 ar630[lin-12::sl2::2xnls-GFP] This study GS9032 lin-1(n304); ar630[lin-12::sl2::2xnls-GFP] This study GS8659 rde-1(ne300); arTi106[lin-31p::rde-1::unc- This study 54 3’UTR]; nre-1(hd20) lin-15b(hd126) GS8804 arSi22;sel-10(ar41) This study

136