Crimpy Sorts a Bmp Into the Regulated Secretory Pathway

CRIMPY SORTS A BMP INTO THE REGULATED SECRETORY PATHWAY

FOR ACTIVITY-DEPENDENT RELEASE IN DROSOPHILA MOTORNEURONS

REBECCA ELLEN JAMES

Submitted in partial fulfillment of the requirements for the

degree of Doctor of Philosophy

Dissertation Advisor: Heather T. Broihier

Department of Neurosciences

CASE WESTERN RESERVE UNIVERSITY

May, 2013

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of

Rebecca E. James

candidate for the doctoral degree*.

Ben Strowbridge

(chair of the committee)

Heather Broihier

Christopher Wilson

Jocelyn McDonald

Stephen Maricich

March 14, 2013

*We also certify that written approval has been obtained for any proprietary material contained herein.

DEDICATION

To my amazing friends and family. Without your encouragement and confidence

in my ability, I would never have dreamt so big, nor would I have had the

confidence and conviction to pursue those dreams.

TABLE OF CONTENTS

List of Tables 4

List of Figures 5

Acknowledgements 7

Abstract 9

Chapter 1: Introduction 13

Summary 14

The Neuromuscular System of Drosophila 15

Synaptogenesis at the Drosophila NMJ 17

Developmental Expansion of the NMJ 18

BMP Signaling 19

Regulation of BMP Signaling 22

Retrograde BMP Signaling During NMJ Development 25

Regulation of BMP Signaling at the Drosophila NMJ 30

Regulated Secretion of Growth Factors 33

Aims of Thesis 36

Figures 1.1-1.3 43

Chapter 2: Crimpy inhibits the BMP homolog Gbb in 49 motorneurons to enable proper growth control at the

Drosophila neuromuscular junction

Abstract 50

Introduction 51

Materials and Methods 55

Results 60

Discussion 73

Figures 2.1 – 2.7 80

Table 2.1 – 2.4 94

Chapter 3: Crimpy sorts a BMP into the regulated secretory 98 pathway for activity-dependent release in Drosophila motorneurons

Abstract 99

Introduction 100

Materials and Methods 104

Results and Discussion 111

Conclusions 125

Figures 3.1-3.6 127

Table 3.1 142

Supplementary Figures 3.1-3.3 143

Chapter 4: General Discussion 150

Challenging the classic BMP signaling paradigm at the 151

Drosophila NMJ

Activity-dependent secretion of motorneuronal Gbb and 153

neurotransmission

Cmpy is a novel sorting receptor for BMP delivery into the 159

RSP

Bibliography 165

LIST OF TABLES

2.1 cmpy loss-of-function and BMP genetic interaction 94

phenotypes at the NMJ

2.2 Gbb and Cmpy gain-of-function phenotypes at the NMJ 95

2.3 Gbb and Cmpy gain-of-function phenotypes in the wing 96

2.4 gbb RNAi phenotypes at the NMJ 97

3.1 Verification of transgene function at the NMJ 142

LIST OF FIGURES

1.1 Organization of the Larval Motor System 43

1.2 Canonical BMP Signal Transduction 45

1.3 BMP Signaling Pathways at the Drosophila NMJ 47

2.1 CG13253 expression analysis and allele generation 80

2.2 cmpy functions in motorneurons to attenuate NMJ expansion 82

2.3 cmpy acts in the BMP signaling pathway upstream of the 84

BMP Type II receptor wit

2.4 Overexpression of cmpy suppresses NMJ expansion in 86

larvae overexpressing neuronal gbb

2.5 Overexpression of cmpy suppresses gbb overexpression 88

phenotypes in the wing disc

2.6 RNAi-mediated knockdown of Gbb in motorneurons 90

suppresses the cmpy LOF phenotype

2.7 Cmpy physically interacts with the Gbb precursor protein 92

3.1 Crimpy does not alter Gbb processing or secretion 127

3.2 Crimpy is necessary and sufficient for Gbb localization at 129

the NMJ

3.3 Crimpy and Gbb colocalize within dense core vesicles of 131

Drosophila motorneurons

3.4 Activity regulates Gbb localization at the NMJ 134

3.5 Synaptic transmission is impaired at cmpy mutant NMJ 137

3.6 Model of Crimpy-dependent Gbb trafficking to the NMJ 139

S3.1 Cmpy-Venus localization at the NMJ 143

S3.2 DVGLUT localization is increased at cmpy∆8 NMJs 146

S3.3 pMad accumulation is increased at cmpyΔ8 NMJs 148

4.1 Proposed Model for Distinguishing BMP Pathways at the 163

NMJ

ACKNOWLEDGEMENTS

First, I would like to thank my thesis advisor, Dr. Heather T. Broihier for her support and encouragement. She provided a positive, comfortable, and compelling training environment. Heather also understood my personal needs for growth as a scientist, and gave me room to explore and to think independently, which I found invaluable.

I also thank the members of my thesis committee, Drs. Stephen Maricich, Ben

Strowbridge, Jocelyn McDonald, and Christopher Wilson for their advice, suggestions, and technical assistance. I am especially thankful to Chris. Without his help and the use of his equipment, I would not have been successful in developing my skills in electrophysiology. Chris had eminent patience with me, and taught me how to really appreciate the days when things work well in a physiology lab.

I am deeply grateful to the other members of the Broihier lab, past and present–

Nan Liu, Crystal Miller, Inna Nechipurenko, Yi-Lan Weng, Chris Dejelo, James

Sears, and Colleen McLaughlin–for their help and encouragement, and for creating a fun lab environment which I gladly considered my home away from home. In particular Nan has been incredible in her technical assistance; she is always willing to help, and always puts forth her best effort. Additionally, she has been my “lab Mom” for the past six years, lending emotional support and feeding me whenever she gets an opportunity, as any great mom would do. I would also

like to acknowledge the Neurosciences Department for promoting collaboration and for providing a great sense of community. It is truly a pleasure to feel confident that I can approach anyone in the Department with a question or an idea and receive a considered response or constructive feedback.

My family has always been my greatest source of support and encouragement.

My loving and father and mother, Dwight and Barbara James, sisters, Melissa

James-Jackson and Sara James, and my best friend since childhood, Christina

Rivera, have always believed in me, even when I have doubted myself. For that I cannot express enough gratitude. My nephews Cole David James and Wesley

Joseph Jackson are a constant source of joy. From my family I have gained a measure of strength and perseverance that I believe can carry me across the finish line of any goal that I set for myself.

Lastly, I could not go without mentioning the climbing community in Northeast

Ohio. I have been so inspired by all of my friends in this wonderful community of ours, especially Chick Holtkamp, Sarah Wilson-Jones, Niki Zmij, Liz Yokum, and

Noah Gostout. The “try hard” that I have learned from each of you as we dance on the rock translates into every aspect of my life. Thanks for helping me become a 5.13 scientist!

Crimpy Sorts a BMP into the Regulated Secretory Pathway for Activity-

Dependent Release in Drosophila Motorneurons

Abstract

REBECCA ELLEN JAMES

Neural circuits integrate experience and store information by the formation and

remodeling of synapses. Defining the signaling pathways that underlie such

plasticity has been a major goal of modern neuroscience. BMP signaling critically

regulates both morphological plasticity and neurotransmission in motorneurons at

the Drosophila neuromuscular junction (NMJ). However the mechanisms that

establish BMP pathway directionality and diversify pathway action in

motorneurons remain elusive. In this thesis, I describe a novel regulator of the

BMP signaling pathway in Drosophila motorneurons that I name Crimpy (Cmpy).

Cmpy acts at the interface between morphological plasticity and synaptic

transmission to specify a role for motorneuronal BMP ligand in promoting

synaptic transmission, as opposed to morphological growth, at the NMJ.

Using the Drosophila NMJ as a model system, I examined the role of Cmpy in

motorneuron development. BMP signaling scales growth of the presynaptic axon arbor to growth of the postsynaptic muscle during larval development in order to preserve synaptic input onto the muscle cell. Genetic analyses suggest that the

relevant BMP ligand at the NMJ, Glass bottom boat (Gbb), acts in a retrograde

fashion, secreted from the postsynaptic muscle to activate presynaptic BMP

receptors and promote motoraxon growth. However establishment of this

directionality is unclear since Gbb is also produced by and active within motorneurons, albeit to regulate synaptic transmission as opposed to morphological growth. I found that loss of cmpy results in excessive presynaptic growth due to ectopic motorneuron-derived, autocrine pro-growth BMP signaling at the NMJ.

Biochemical analyses demonstrated that Cmpy binds to precursor BMP.

Expressing a novel Gbb-HA transgene within larval motorneurons that preserves

protein processing and secretion, and functions like a wild-type Gbb transgene, I

detected presynaptic Gbb-HA localization at the NMJ. Cmpy is necessary and

sufficient for Gbb-HA localization at the NMJ. Gbb-HA and Cmpy-Venus

transgenes colocalize at presynaptic terminals and also within motor nerves,

where they are observed in discrete puncta reminiscent of trafficking vesicles.

Furthermore, Gbb-HA and Cmpy-Venus both colocalize with a marker for dense

core vesicles (DCVs) of the regulated secretory pathway (RSP) at the NMJ and in individual puncta within motor nerves, which indicated that Gbb is secreted from DCVs in response to synaptic activity. Indeed, by stimulating synaptic activity in motorneurons using two independent measures, high K+ depolarization

and by activating the blue light-gated cation channel, channelrhodopsin, in

motorneurons, we detect activity-dependent changes in Gbb-HA levels at the larval NMJ.

Growth factors of the neurotrophin and TGFβ superfamilies are secreted from vertebrate neurons in response to synaptic activity in addition to constitutive secretion. Sorting receptors for the neurotrophin BDNF have been identified that interact with precursor BDNF and direct its delivery into the RSP. This reminded us of interactions between Cmpy and precursor Gbb. Additionally, activity- dependent growth factor secretion promotes synaptic transmission at vertebrate synapses. Similarly, neuronal Gbb is critical for proper neurotransmission at the

Drosophila NMJ. We hypothesized that Cmpy is a sorting receptor that delivers

Gbb into the RSP for activity-dependent secretion in larval motorneurons, and that activity-dependent Gbb signaling strengthens synaptic transmission. In line with this hypothesis, I found that Cmpy is necessary for activity-dependent changes in Gbb-HA levels at the NMJ. Also, evoked synaptic transmission is impaired at cmpy mutant NMJs, despite overgrowth of cmpy mutant terminals.

Taken together, my thesis work identifies a novel pathway that diversifies BMP pathway action in larval motorneurons by sorting the BMP ligand into DCVs of the RSP for activity-dependent secretion. We propose that sorting of Gbb into the

RSP in Drosophila motorneurons defines its role in synaptic transmission, as opposed to growth, at the NMJ. To date, Cmpy is the first BMP sorting receptor identified in any species. Given the high degree of conservation of key

developmental signaling pathways between vertebrates and invertebrates, it is likely that vertebrate homologs of Cmpy similarly sort TGFβ ligands into the RSP for activity-dependent secretion from vertebrate neurons.

CHAPTER 1: Introduction

Summary

Wiring a nervous system requires the directed flow of information. Electrical information flow is an essential means of communication between neurons, and is driven by action potential propagation and neurotransmitter release from the presynaptic compartment to activate neurotransmitter receptors on the postsynaptic cell. In addition to anterograde communication via neurotransmitter release at the synapse, signaling pathways initiated by postsynaptic cells modulate presynaptic input in a retrograde fashion. Presynaptic neurons can also integrate information about the relative activity or strength of an individual synapse by signaling in an autocrine fashion in response to synaptic activity.

Whereas the direction of neurotransmission is established by the inherent difference in distribution of neurotransmitter-containing synaptic vesicles and receptors in the pre- and postsynaptic compartments, respectively, establishing directionality of signaling pathways that modulate synaptic strength is more complex. Signaling pathway flow can be further complicated if the same pathway is used at a particular synapse to achieve distinct cellular outcomes, like modulating synaptic strength versus synaptic morphology. Clarifying the mechanisms set by neurons to establish signaling pathway direction could provide a means to inhibit or enhance one pathway without disrupting another.

This could be useful in developing specific therapies to treat signaling pathway dysfunction that alters neurotransmission and leads to neurological disorders.

Since there is a high degree of signaling pathway conservation between

vertebrates and invertebrates, it is reasonable to predict that characterizing the

mechanisms that confer pathway directionality in more simple, genetically

tractable systems like Drosophila will shed light onto pathway directionality in

vertebrates.

The Neuromuscular System of Drosophila

For decades, developmental neuroscientists have employed both vertebrate and

invertebrate models to unearth cues that guide developing axonal and dendritic

processes along their specified trajectories and their integration into functional

neural circuits. The Drosophila larval neuromuscular junction (NMJ) has emerged

as a favorite model for investigating synapse formation, given the accessibility of

singly identifiable, highly stereotyped synaptic terminals, and the powerful

techniques available to probe both structural and functional questions at this

synapse (Collins and DiAntonio 2007). The Drosophila NMJ is glutamatergic, in contrast to the cholinergic NMJ of vertebrates, imparting more similarity to excitatory central synapses of vertebrates than the vertebrate NMJ. Furthermore, many molecules that direct synaptic differentiation and function at the larval NMJ are highly conserved in vertebrates, making it likely that elucidation of gene function in Drosophila will lend insight into excitatory synapse development in the vertebrate CNS (Keshishian et al. 1996, Collins and DiAntonio 2007).

During late stages of Drosophila embryogenesis, efferent motorneuron fibers project from the central nervous system (CNS) and navigate through the

periphery, responding to both attractive and repulsive cues, until their appropriate

muscle target is reached. Upon target recognition, the formation of functional

synapses commences. Taking advantage of the stereotypy of this system and

the power of modern Drosophila genetics, a multitude of signaling molecules

have been uncovered that regulate guidance, synaptogenesis, and synaptic

expansion. These molecules largely belong to well-characterized, evolutionarily

conserved signaling pathways, like the bone morphogenetic protein (BMP) and

Wingless/Wnt pathways, which drive multiple developmental processes in all

species.

The motor system of Drosophila consists of 36 unipolar motorneurons per

hemisegment of the ventral ganglion, comparable to the vertebrate spinal cord,

that project axons out of the CNS to innervate 30 muscles in each hemisegment of the body wall (Landgraf and Thor 2006). Motorneuron cell bodies reside in the cortex of the ventral ganglion and project into the neuropil, where they extend dendrites in the synaptic region of the neuropil and send an axon into the periphery (Fig. 1.1A). Axons project either ipsi- or contralaterally to exit the CNS through one of three nerves: the intersegmental nerve (ISN, Fig. 1.1B), the segmental nerve (SN), or the transverse nerve (TN) (Landgraf and Thor 2006,

Sanchez-Soriano et al. 2007).

Once in the periphery, motoraxons respond to both short- and long-range guidance cues to navigate to their proper muscle targets. Short-range signaling

involves contact-mediated attraction and repulsion, whereas long-range guidance is mediated by diffusible, target-derived factors (Tessier-Lavigne and Goodman

1996). Upon reaching a potential target, a combinatorial assessment of attractive and repulsive cues, mostly cell-adhesion molecules, determines the pairing of pre- and postsynaptic partners (Jin 2002).

Synaptogenesis at the Drosophila NMJ

Initial target contact drives functional synapse assembly through concerted

differentiation of the pre- and postsynaptic compartments. At the Drosophila

NMJ, a functional synapse is composed of glutamate receptor clusters within the

postsynaptic density apposing presynaptic glutamate release sites, or active

zones (AZ) (Marques 2005). The Drosophila homolog of vertebrate AZ structural protein complex ELKS/CAST, Bruchpilot (BRP), clusters Ca2+ channels and

promotes SV release (Kittel et al. 2006, Fouquet et al. 2009). Presynaptic activity

promotes the clustering of postsynaptic glutamate receptors, yet the details of

this process remain controversial (Broadie and Bate 1993, Saitoe et al. 2001,

Daniels et al. 2006). Specifically, whether vesicular or non-vesicular glutamate

drives the clustering of receptors is under debate. Within an hour of initial target contact, activity can be detected from immature embryonic NMJs (Featherstone and Broadie 2000).

Developmental Expansion of the NMJ

The foundations for a functional motor system are laid by the culmination of

these synaptogenic events by the time of larval hatching at 24 hours after egg- laying, after which muscle contraction and motion can be achieved. Larvae grow dramatically prior to pupariation and metamorphosis, experiencing a 100-fold increase in muscle area during larval development (Atwood et al. 1993,

Keshishian et al. 1993). To accommodate the increase in muscle area, motorneurons elaborate their axon terminals by adding up to 10-fold more synaptic boutons and AZs, thus preserving synaptic strength as larvae grow

(Schuster et al. 1996). This process is termed developmental presynaptic

expansion, and is distinct from embryonic synaptogenesis (Packard et al. 2003).

The Wingless/Wnt signaling pathway modulates both presynaptic expansion and

postsynaptic differentiation. Wingless is secreted by motorneuron terminals at the

NMJ, and signals presynaptically through a canonical local pathway to regulate

bouton number and microtubule dynamics (Miech et al. 2008). Postsynaptically,

Wingless binds to its receptor DFrizzled2, the complex is endocytosed and the C-

terminus of the receptor is cleaved. After cleavage the C-terminus translocates to

muscle nuclei where it is thought to regulate a synaptic growth promoting

program transcriptionally (Mathew et al. 2005). In the absence of Wingless, NMJ

growth is impaired, boutons exhibit altered AZ ultrastructure, and postsynaptic

glutamate receptor distribution and subsynaptic reticulum development are

perturbed (Packard et al. 2002). In addition to anterograde/autocrine Wingless

signaling at the NMJ, BMP signaling was also identified as a prominent regulator of developmental axon expansion at the Drosophila NMJ.

BMP Signaling

BMPs are members of the evolutionarily conserved TGFβ superfamily of secreted growth factors. They participate in a multitude of developmental and homeostatic processes, and mutations in BMP signaling components lead to human disease and cancer (Attisano and Wrana 2002). BMPs can act as classic morphogens, establishing gradients that signal at long-range distances from the signal-sending cell and elicit distinct outcomes based upon the concentration of available ligand. BMP signaling has been well studied in several systems, including mouse, chick, and Drosophila, and regulates cell fate, proliferation, neural differentiation, organ patterning and growth control, and cell death, among

other processes (Mishina 2003, Liu and Niswander 2005, Affolter and Basler

2007, Miyazono et al. 2010). Alternatively, BMPs can function as short-range

local cues, as in germline stem cell niches in the Drosophila ovary and testis,

where they maintain stem cell self-renewal (Ma and Xie 2011). Many extracellular factors have been identified that regulate the establishment and maintenance of BMP gradients (O'Connor et al. 2006, Bier 2008).

Traditionally, BMP ligands are thought to be composed of disulfide-linked dimeric proteins, originally synthesized as large precursors and then proteolytically cleaved by furin convertases to release C-terminal, mature peptides of

approximately 100–140 amino acids in length (Kawabata et al. 1998). Recently however, a large biologically active BMP was isolated in Drosophila that can act at a longer range than the smaller, classic peptide. A novel, evolutionarily conserved furin cleavage site in Glass bottom boat (Gbb), the Drosophila

BMP5/6/7 ortholog, generates the larger BMP product of 338 amino acids.

Mutations in this cleavage site lead to developmental abnormalities in humans, underlining the importance of this form of processing (Akiyama et al. 2012).

Whether the larger product forms heterodimers with the small peptide, and how it interacts with BMP receptors to mediate downstream signal transduction, remains open for investigation.

In classic BMP signal transduction, binding of homo- or heterodimers of processed, mature peptide induces the association of type I and type II serine- threonine kinase receptors (Fig. 1.2). Both receptors are single-pass transmembrane proteins that exist as dimers on the cell-surface in the absence of ligand, with intracellular kinase domains. Unlike TGFβs, which bind specifically to type II receptors, BMPs can bind to type I receptors in the absence of type II receptors, but the association is greater in the presence of both (Miyazono et al.

2010). Upon ligand binding and receptor association, type II receptors phosphorylate, and thereby activate, the type I receptors. Phosphorylated type I receptors then phosphorylate a receptor-associated Smad (R-Smad) multimer, augmenting its affinity for Co-Smads and unveiling its nuclear import signal (Shi and Massague 2003). R-Smad/Co-Smad complexes translocate to the cell

nucleus, and associate with coactivators and corepressors to mediate changes in gene transcription.

BMP signaling is pivotal to development of the nervous system in both

vertebrates and invertebrates. In vertebrates, inhibition of BMP signaling is

crucial for induction of the neural plate from non-neural ectoderm. At later stages,

BMP signaling directs formation of the neural crest and patterns the dorsal spinal

cord, where BMP signaling is central for dorsal cell fate specification (Liu and

Niswander 2005). In the absence of BMP signaling, dorsal-most interneuron

populations are lost and more ventral interneuron populations expand into the

dorsal domain of the spinal cord (Chesnutt et al. 2004). BMP signaling also

specifies cerebellar granule cell fate (Alder et al. 1999) and directs dendrite

morphogenesis of cultured cerebellar cortical, hippocampal, and sympathetic

neurons (Lein et al. 1995, Le Roux et al. 1999, Withers et al. 2000).

As in vertebrates, BMP signaling is vital during early embryonic patterning in

Drosophila. The BMP2/4 ortholog, Decapentaplegic (Dpp), acts as a classic

morphogen to specify dorsal cell fates during early embryogenesis. Loss of Dpp

ventralizes embryos, such that ventral cell fates are adopted by tissues that

would otherwise acquire dorsal phenotypes (Arora et al. 1996). Further, Dpp

specifies mesodermal tissues of the gut and heart during early embryonic

development (Frasch 1995), and regulates the proliferation and differentiation of

cells in the imaginal discs, structures that give rise to adult organs (Arora et al.

1996). Importantly, as mentioned previously, BMP signaling is essential for regulating motorneuron morphology and function at the larval NMJ.

Regulation of BMP Signaling

Given the diverse and critical roles of BMP signaling during development, it is easy to predict multiple levels of regulation that preserve temporal and spatial signaling resolution. Both extracellular and intracellular modes of BMP regulation have been identified that influence gradient establishment and ligand processing/secretion, respectively, although mechanisms of extracellular BMP regulation are better understood. Many of the known extracellular regulators of

BMP signaling are evolutionarily conserved, soluble proteins that directly bind to

BMPs through cysteine-rich repeat (CRR) domains, and either promote or antagonize BMP activity in a context-dependent manner (Umulis et al. 2009,

Zakin and De Robertis 2010).

Several secreted BMP regulators, such as Chordin/Short gastrulation (Sog),

Twisted gastrulation (Tsg), Noggin, and Follistatin, inhibit signaling by binding to

BMPs and sequestering them from receptors (Zakin and De Robertis 2010).

Molecules that sequester BMPs can also promote signaling by facilitating their movement through the extracellular milieu (Umulis et al. 2009). In Drosophila,

Sog and Tsg facilitate transport of the Dpp to the dorsal midline during embryonic development (Shimmi et al. 2005). Chordin and Tsg similarly complex with BMPs to regulate their flow during dorsal-ventral patterning of developing Xenopus

embryos (De Robertis 2009). The extracellular matrix molecule collagen IV

promotes BMP-Sog-Tsg complex flow in Drosophila, and this function is likely

conserved in vertebrates (Wang et al. 2008). Once the complexes reach their

proper signaling niches, local Tolloid metalloproteinases digest Chordin/Sog,

releasing the BMP and allowing receptor activation (Piccolo et al. 1997, Shimmi

et al. 2005, Zakin and De Robertis 2010). Thus, extracellular BMP antagonists

can provide temporal regulation and promote BMP activity by preventing

premature ligand endocytosis and degradation.

Another secreted BMP-binding protein, Crossveinless-2 (CV2), also interacts

with BMP-Chordin/Sog-Tsg complexes. Complexes that come into contact with

areas rich in CV2, which is often tethered to cell membranes by interactions with

membrane-bound heparan sulfate proteoglycans and binds BMP-Chordin/Sog

complexes with high affinity, can be concentrated (Ambrosio et al. 2008, Serpe et

al. 2008, Zakin and De Robertis 2010). In addition to binding BMP-Chordin/Sog-

Tsg complexes, CV2 interacts with the type I BMP receptor Thickveins in

Drosophila. In low levels of CV2, an exchange of BMP between CV2 and

Thickveins can occur, whereas high levels of CV2 inhibit signaling (Serpe et al.

2008). The pro-BMP function of CV2 has been proposed to lie in its ability to act as a molecular sink, amassing BMP-Chordin/Sog-Tsg complexes near receptors, where Tolloid can cleave Chordin/Sog to release high concentrations of BMP or

CV2 can directly pass the ligand to the receptor (Zakin and De Robertis 2010).

In addition to regulating BMP flow and gradient dynamics, several BMP interacting proteins have been identified that modulate ligand processing and/or secretion in signal-sending cells. As mentioned earlier, BMPs are synthesized as large precursors which undergo cleavage by furin convertases in the secretory pathway to produce smaller, “mature” signaling peptides, although it is now appreciated that larger BMP products are biologically active (Akiyama et al.

2012). Vertebrate cysteine-rich in motorneurons-1 (CRIM-1) binds to BMPs through a CRR domain like Chordin, and presumably does so early in the secretory pathway to prevent ligand processing and secretion in vitro (Wilkinson et al. 2003). CRIM-1 may also regulate specification of ventral motorneuron and interneuron cells types in the developing chick spinal cord (Kolle et al. 2003).

Members of the DAN family of secreted glycoproteins, including Dan, Cerberus,

Gremlin, and Sclerostin, can bind to and inhibit BMPs via extracellular sequestration (Gazzerro and Canalis 2006). However, for both Sclerostin and

Gremlin, intracellular modes of inhibition have also been established. In the developing mouse lung, Gremlin can bind to the precursor of BMP4, and interactions between Gremlin and BMP4 in vitro prevent secretion of active BMP

(Sun et al. 2006). Interestingly, this interaction was mapped to an arginine-lysine rich region in the Gremlin DAN domain, not its CRR domain. Similarly, Sclerostin interacts with precursor BMP7 in vitro, preventing secretion and promoting its proteasomal degradation (Krause et al. 2010).

These studies touch on BMP regulation at the level of ligand availability prior to receptor activation and emphasize the complexity of regulatory modes in place that fine-tune BMP signal transduction. Numerous pathways also control the timing and intracellular localization of signaling components downstream of ligand-receptor interactions. Given the recent identification of intracellular inhibition via DAN family members, it is likely that future studies will continue to unveil novel mechanisms that impact BMP processing, secretion, and availability.

Retrograde BMP Signaling During NMJ Development

In 2002, back to back publications in Neuron converged on a role for the BMP type II receptor Wishful thinking (Wit) in regulating morphology and synaptic transmission at the Drosophila NMJ. These studies provided compelling evidence that Wit is expressed by specific subsets of neurons in developing embryos and larvae, and that in its absence developmental axon expansion is impaired, evidenced by approximately 50% fewer synaptic boutons at the NMJ

(Aberle et al. 2002, Marques et al. 2002). Synaptic transmission is also compromised at wit mutant NMJs: evoked synaptic release, quantal content, and spontaneous release frequency are all significantly decreased. Additionally, NMJ ultrastructure is perturbed. Electron micrographs revealed detachment of the presynaptic membrane from the postsynaptic membrane, primarily near active zones (Aberle et al. 2002, Marques et al. 2002). In wit mutant embryos, accumulation of the phosphorylated R-Smad, Mothers against Dpp (Mad), in motorneuron nuclei was completely abolished (Marques et al. 2002). All aspects

of the wit loss-of-function (LOF) phenotype are rescued by neuronal expression

of a wild-type Wit transgene, demonstrating that wit functions presynaptically at the NMJ.

After the identification of Wit as a principle regulator of NMJ morphology and function, the key players that transduce the BMP signal at the NMJ were pieced together (Fig. 1.2). Loss of Mad, the co-Smad, Medea, and the type I receptors

Saxophone (Sax) and Thickveins (Tkv) phenocopies NMJ undergrowth and defects in synaptic transmission observed at wit LOF NMJs (Rawson et al. 2003,

McCabe et al. 2004). Loss of the BMP ligand Gbb, boasts comparable NMJ

growth and synaptic transmission defects (McCabe et al. 2003). In this study, embryonic in situ hybridization revealed gbb RNA expression in both the CNS

and the body wall musculature. However, the postsynaptic muscle was assumed

to be the primary source of the BMP ligand that regulates NMJ development based upon rescue experiments. Postsynaptic expression of Gbb alone completely restored NMJ arbor size in a strong hypomorphic background, whereas neuronal expression only partially rescued growth (McCabe et al. 2003).

This led to the idea that Gbb is a retrograde ligand at the NMJ, secreted from the

muscle and exerting its action on presynaptic motorneurons, to instruct axon

expansion and couple motoraxon growth to muscle growth. However, if solely

postsynaptic Gbb is responsible for motorneuron differentiation, why is Gbb also

expressed in the CNS?

Insight into this discrepancy can be gleaned from synaptic transmission rescue

experiments. Postsynaptic expression of Gbb in a strong gbb hypomorphic background partially rescued synaptic function, whereas panneuronal expression completely restored defects in transmission (McCabe et al. 2003). A later study examining synaptic transmission in a true gbb null background, however, found that postsynaptic expression of ligand was insufficient to confer even partial rescue of synaptic transmission (Goold and Davis 2007). In this study, pre- and

postsynaptic expression of ligand comparably rescued NMJ morphology,

whereas only panneuronal expression of ligand significantly restored function.

These studies implicate neuronal Gbb ligand as a key modulator of basal

synaptic transmission.

In accordance with a role for motorneuronal BMP ligand in regulating synaptic

transmission, retrograde BMP signaling between postsynaptic motorneurons and

presynaptic interneurons in the CNS strengthens transmission at this central

synapse (Baines 2004). In a gbb null background, evoked synaptic transmission

is impaired between presynaptic interneurons and postsynaptic motorneurons.

Furthermore, overexpression of wild-type Gbb in postsynaptic motorneurons is

sufficient to strengthen the evoked response from presynaptic interneurons

(Baines 2004). This study highlights the ability of motorneuron-derived BMP to

regulate synaptic transmission, at least at this central synapse.

In addition to modulating basal synaptic function, Gbb-mediated BMP signaling

through Wit and Mad is necessary to express synaptic homeostasis, a

compensatory increase in quantal content when postsynaptic GluRs are blocked

by the use-dependent glutamate receptor antagonist, philanthotoxin (Goold and

Davis 2007). Synaptic homeostasis is rescued by either pre- or postsynaptic

expression of Gbb ligand in a true gbb null background. However as previously mentioned, only presynaptic expression of BMP is capable of significantly

rescuing defects in evoked synaptic transmission in the gbb null background.

Again this strengthens the idea that neuronal and muscle-derived pools of Gbb

signal distinct outcomes at the NMJ.

Intriguingly, the same study revealed a persistent developmental requirement for

BMP signaling to regulate morphological NMJ expansion, neurotransmission,

and to confer the competency to express synaptic homeostasis. Inhibiting BMP

signaling at various time points during larval development revealed grades of

impairment in each of these processes, the severity of defects mirroring the

duration of BMP signaling inhibition (Goold and Davis 2007). Of note, inhibiting

BMP signaling late during larval development—at a time when NMJ expansion is

not impaired, likely reflecting the end of larval growth prior to pupariation—still

resulted in defects in evoked synaptic transmission and synaptic homeostasis.

This demonstrates a continual need for BMP signaling at the NMJ to maintain

synaptic function at a time when presumably the muscle-derived, pro-growth pool of BMP ligand would no longer be active.

Stressing the complexity of divergent BMP signaling pathway action in motorneurons, yet another role can be ascribed to Gbb and Wit in regulating morphology and function at the NMJ. Gbb signaling through Wit activates a presynaptic non-canonical BMP signal transduction cascade within motorneurons that controls synaptic stability (Eaton and Davis 2005). This study found that canonical BMP signal transduction contributes to synapse stability, however loss of wit results in a dramatic increase in synapse destabilization compared to loss of the other components in the canonical pathway, indicating that canonical signaling alone cannot account for impaired synapse stabilization in wit mutants.

The Ser/Thr kinase LIM Kinase-1 (DLIMK-1) was found to promote synapse stability downstream of Wit through a non-canonical BMP pathway at the NMJ

(Eaton and Davis 2005). Significantly, neuronal expression of wild-type DLIMK-1 in a wit mutant background restored defects in synaptic transmission at wit LOF

NMJs (Eaton and Davis 2005), hinting to a role for non-canonical BMP signaling in modulation of synaptic transmission at the NMJ.

Together, these data strongly imply that muscle is not the only source of active

BMP ligand regulating motorneuron differentiation and function at the NMJ.

Further, these studies emphasize the various functions of BMP signaling in motorneurons, and point to a broad role for neuronal BMP in regulating synaptic transmission that is separable from pro-growth, presumably muscle-derived BMP signaling. I propose that at least two distinct BMP signaling pathways are active at the NMJ: retrograde BMP signaling from the muscle scales motorneuron

growth to muscle growth during the course of larval development, whereas

autocrine BMP signaling from motorneurons regulates synaptic transmission

(Fig. 1.3). That BMP signaling can produce such diverse responses at a particular location hints to obscured levels of signaling regulation in motorneurons to distinguish motorneuron-derived BMP from muscle-derived

BMP.

Regulation of BMP Signaling at the Drosophila NMJ

At the Drosophila NMJ, the BMP regulatory events described to date primarily modulate BMP signaling downstream of receptor activation. Presynaptically, endosomal/lysosomal sorting is critical for attenuating BMP signaling. Spinster

(Spin), Nervous wreck (Nwk), Spichthyin (Spict), and Endosomal maturation defective (Ema) are all thought to diminish BMP signaling by promoting internalization of BMP receptors in motorneurons (Sweeney and Davis 2002,

Wang et al. 2007, O'Connor-Giles et al. 2008, Kim et al. 2010). Loss of any of these results in significant increases in the number of synaptic boutons at the

NMJ, and for Nwk, Spict, and Ema, an increase in the accumulation of local phosphorylated R-Smad, Mad (pMad) at the NMJ. The correlation between increased pMad at the NMJ and morphological overgrowth suggests a local role for BMP signaling in modulating growth, in addition to the canonical nuclear pathway revealed by mutations in wit (Marques et al. 2002). Interestingly, despite dramatic morphological overgrowth and increased BMP signaling, as indicated by increased pMad accumulation at the NMJ, spict and nwk mutant NMJs have

impaired evoked synaptic transmission (Sweeney and Davis 2002, Coyle et al.

2004).

In addition to regulating receptor localization, molecules have also been identified

that direct localization of the downstream effector, Mad. Nemo (Nmo) kinase

phosphorylates Mad at a site distinct from type I BMP receptor phosphorylation,

and this modification is necessary for proper pMad localization in motorneurons

(Merino et al. 2009). In the absence of Nmo phosphorylation, pMad accumulates

at the NMJ, with a corresponding decrease in nuclear pMad, and terminal growth

is impaired. This indicates that while local pMad may be sufficient to drive NMJ

overgrowth in backgrounds like spict or nwk with increased BMP signaling at the

NMJ, local BMP signaling is insufficient to drive growth in the absence of appropriate canonical nuclear signaling. Despite a decrease in the number of synaptic boutons in nmo mutants, however, synaptic transmission is unaffected

(Merino et al. 2009). Conversely, Importin-β11 (Impβ11), a member of the importin nuclear import receptor family, promotes axon growth by retaining pMad

at the NMJ. In impβ11 mutants, pMad fails to accumulate at the NMJ while

nuclear pMad is unaffected, there are significantly fewer synaptic boutons, and

evoked synaptic transmission is impaired (Higashi-Kovtun et al. 2010). These

studies emphasize the importance of tightly controlling the levels of local versus

nuclear pMad in control of NMJ differentiation. Further studies are needed to

clarify the specific roles of nuclear versus local pMad in synaptic growth and

function.

Beyond these examples of BMP pathway regulation at the level of pathway

component localization, little is known about modulating BMP signaling prior to

ligand-receptor interactions. On the postsynaptic side of the synapse, a putatively

autocrine TGFβ pathway activates canonical TGFβ signaling in the muscle to modulate expression of the relevant BMP ligand, Gbb. In mutants of the TGFβ

ligand, Dawdle (Daw), developmental NMJ expansion is impaired and

postsynaptic expression of Gbb in larval muscles is reduced, indicating daw is necessary for Gbb expression (Ellis et al. 2010). However, overexpression of

Daw does not enhance synaptic growth or Gbb expression, indicating that levels of Daw are not instructive for Gbb-mediated NMJ expansion per se.

Alternatively, other mechanisms may ensure fidelity of postsynaptic Gbb

expression and/or secretion. Indeed, postsynaptic Cdc42-interacting protein-4

(dCIP4) restrains NMJ growth by inhibiting postsynaptic Gbb secretion (Nahm et al. 2010), whereas the postsynaptic Cdc42-selective guanosine triphosphatase-

activating protein, dRich, promotes growth by stimulating Gbb release from the

postsynaptic muscle (Nahm et al. 2010).

Each step of regulating BMP signal transduction—whether at the level of

receptor availability, downstream effector localization, or ligand secretion—is

complicated. The source, directionality, and canonical nature of BMP signal

transduction in Drosophila motorneurons has largely been taken for granted. The

aforementioned studies raise several important questions regarding the source(s)

and function(s) of Gbb at the NMJ. If muscle-derived ligand orchestrates morphological growth, and neuronal ligand neurotransmission, how does the motorneuron differentiate BMP signaling to achieve distinct cellular responses at the NMJ? Or, how does the motorneuron perceive muscle-derived ligand as pro- growth, and neuronal ligand as pro-function? How is BMP secretion regulated in motorneurons to promote neurotransmission as opposed to growth at the NMJ?

Regulated Secretion of Growth Factors

Proteins are secreted from neurons through either the constitutive secretory pathway (CSP) or the regulated secretory pathway (RSP). A major entry point into either involves recognition of hydrophobic or transmembrane signal sequences by the signal recognition particle and cotranslational insertion of secreted or plasma membrane bound proteins into the endoplasmic reticulum

(Barlowe and Miller 2013). Upon insertion, the signal sequence may be removed and nascent peptides may be posttranslationally modified to enhance solubility and aid in protein folding (Barlowe and Miller 2013). Secretory cargo is transported to the Golgi complex, where further posttranslational modifications, such as proteolytic processing of peptide precursors, occur as the proteins navigate through the Golgi cisternae (Vazquez-Martinez et al. 2012).

Of the molecules that undergo regulated secretion at synapses, neurotransmitters are secreted from small, clear synaptic vesicles, whereas neuropeptides and growth factors are packaged into dense core vesicles (DCVs)

of the RSP. Proteins are directed into DCVs in the distal-most cisternae of the

Golgi complex, the trans-Golgi network (TGN), by interactions with sorting

receptors (Sossin et al. 1990). Secretory vesicles are actively transported to

release sites by microtubule associated Kinesin motors, and are docked in the

cytoplasm until stimulated to secrete their contents (Burgess and Kelly 1987,

Vazquez-Martinez et al. 2012). In the case of neurons, regulated secretion is

driven by changes in intracellular Ca2+ following neuronal activity. When sorting

of regulated cargo into DCVs is disrupted, regulated cargo is diverted into the

CSP (Burgess and Kelly 1987, Vazquez-Martinez et al. 2012). In the CSP,

vesicles containing integral membrane proteins and proteins destined for

constitutive secretion are directly transported to the plasma membrane and fuse

without stimulation (Vazquez-Martinez et al. 2012).

Growth factors of the neurotrophin family can undergo either constitutive or

regulated secretion in neurons. Of the four neurotrophins, precursors of nerve

growth factor (NGF), neurotrophin-3 and neurotrophin-4/5 are primarily sorted

into the CSP (Thomas and Davies 2005), although NGF also undergoes

regulated secretion in hippocampal slices and primary neuronal cultures

(Thoenen 1995). Brain-derived neurotrophic factor (BDNF), on the other hand, is

preferentially sorted into the RSP, and activity-dependent secretion of BDNF

modulates synaptic plasticity, playing critical roles in long term potentiation and

memory consolidation (Lu 2003, Thomas and Davies 2005, Bekinschtein et al.

2008). Initially thought to localize in postsynaptic dendrites and regulate synaptic

plasticity in a retrograde manner, BDNF and its cleaved propeptide localize presynaptically in large DCVs of excitatory terminals (Dieni et al. 2012).

Interactions between the membrane-associated sorting receptors Sortilin and

carboxypeptidase E (CPE) and precursor BDNF direct proBDNF into the RSP in

hippocampal and cortical neurons, respectively (Chen et al. 2005, Lou et al.

2005).

Like neurotrophins, in addition to constitutive secretion TGFβs can also be

released in an activity-dependent manner. TGF-β2 colocalizes with secretory

granules and is secreted from primary hippocampal neuron cultures following

high potassium depolarization (Specht et al. 2003). Additionally, TGFβ secretion

from a primary hippocampal network in vitro is sufficient to activate downstream

TGFβ signaling within the hippocampal neurons themselves, indicating that activity-dependent TGFβ signaling may act in an autocrine fashion (Lacmann et

al. 2007). The in vivo actions of activity-dependent TGFβ signaling have yet to be characterized. Functional analysis of activity-dependent growth factor signaling in vivo is complicated by the complexity and density of vertebrate central synapses.

Again reflecting neurotrophic growth factor signaling, TGFβs modulate synaptic transmission. Bath application of TGF-β2 enhances evoked postsynaptic currents and increases the amplitude and frequency of miniature spontaneous currents in mouse hippocampal neurons in vitro (Krieglstein et al. 2011). Interestingly, loss of the BMP antagonist Chordin increases presynaptic neurotransmitter release in

the mouse hippocampus, yielding enhanced learning capacity in the Morris water maze (Sun et al. 2007). Further, acute application of BMP6 to hippocampal slices increased release probability and paired-pulse facilitation, indicating that BMP signaling affects transmission presynaptically (Sun et al. 2007). Since BMP6 is primarily expressed by neurons in the hippocampus (Tomizawa et al. 1995), these data suggest that BMP6 promotes synaptic transmission through an autocrine mechanism.

The parallels between BDNF and TGFβ growth factor signaling in undergoing regulated secretion and modulating synaptic transmission begs the question: does autocrine, activity-dependent TGFβ signaling regulate neurotransmission in vivo? As discussed previously, neuronal BMP signaling regulates synaptic transmission at the larval NMJ, however the mechanisms distinguishing pro- function versus pro-growth BMP signaling have not been explored. An appealing hypothesis that will be examined in this thesis is that ligand secretion and timing of motorneuron-derived, pro-neurotransmission BMP signaling is linked to and regulated by activity, thus differentiating it from muscle-derived, pro-growth signaling at the NMJ.

Aims of Thesis

In my thesis work I sought to understand how a single BMP ligand could be utilized by two different cell types to elicit discrete outcomes at a common cellular location. While previous studies suggested that muscle was the predominant

source for the pro-growth BMP ligand Gbb at the NMJ (McCabe et al. 2003),

several studies converged on roles for neuronal Gbb in regulating synaptic

transmission (McCabe et al. 2003, Baines 2004, Goold and Davis 2007). This

implied that motorneurons have a method in place to either alter motorneuronal

BMP appearance or to respond to motorneuronal, pro-neurotransmission BMP differently than muscle-derived, pro-growth BMP.

One could imagine several ways to distinguish between the two pools of BMP ligand at the NMJ. The motorneuron could process or posttranslationally modify

Gbb so that it does not look like muscle-derived ligand. Altering the appearance

of neuronal ligand could allow it to bind a unique composition of receptors at the

NMJ to promote signaling through a divergent BMP pathway, like the non-

canonical DLIMK-1 pathway, whereas retrograde signaling could be transduced

in motorneurons by the canonical BMP cascade. Instead, the motorneuronal pool

could be excluded from the NMJ by sorting receptors in the cell body, preventing

BMP trafficking to presynaptic terminals and thus allowing reception of only the muscle-derived pro-growth ligand in the periphery. Alternatively, the timing of

motorneuronal BMP signaling could be uncoupled from postsynaptic signaling to

bring about a unique intracellular response to motorneuron-derived ligand. For

instance, if neuronal BMP signaling was coupled to a change in the intracellular

environment of the presynapse, like changes in intracellular Ca2+ levels caused

by synaptic activity, downstream signaling effectors could interact with Ca2+-

dependent molecules to distinguish the signaling cascade. Thus, linking

motorneuronal BMP signaling to activity could diversify BMP signaling output at the NMJ. And importantly, these possibilities are not mutually exclusive.

In an effort to characterize a putative ortholog of the vertebrate BMP antagonist

CRIM-1—later named Crimpy (Cmpy)—in Drosophila motorneuron development,

I uncovered a role for Cmpy in restricting larval NMJ growth. Cmpy is a single- pass transmembrane protein expressed in presynaptic motorneurons, but not in postsynaptic muscle. To determine the function of Cmpy, I generated a LOF allele and analyzed the phenotypic consequences of cmpy loss in embryos, which were grossly unaffected, and in larvae. At the larval NMJ, cmpy LOF resulted in NMJ arbor expansion. This overgrowth was reminiscent of overgrowth in mutants for the endocytic adaptor protein, Nwk, which have enhanced BMP signaling (O'Connor-Giles et al. 2008). I carried out genetic interaction experiments and found that overgrowth at cmpy mutants NMJs is dependent on levels of BMP pathway components, indicating that overgrowth is caused by excessive BMP signaling.

I initially hypothesized that Cmpy functions as a cell-autonomous inhibitor of BMP secretion in motorneurons. Misexpression of Cmpy in postsynaptic muscle impaired NMJ growth, supporting the idea that Cmpy cell-autonomously inhibits secretion of BMP. Furthermore, overexpression of the BMP ligand Gbb in motorneurons was sufficient to drive overgrowth at the NMJ, a phenotype that was completely reversed by co-overexpression of Cmpy with Gbb. These

experiments pointed towards a cell-autonomous mechanism to restrict

motorneuron-derived pro-growth BMP signaling. Importantly, knocking down Gbb

in motorneurons partially suppressed NMJ overgrowth in cmpy mutants, further

substantiating the hypothesis that Cmpy restrains autocrine, pro-growth BMP

signaling in motorneurons. Yeast two-hybrid and immunoprecipitation

experiments revealed that Cmpy binds specifically to precursor Gbb, indicating

regulation of BMP signaling early in the secretory pathway.

To investigate the mechanism behind Cmpy-dependent regulation of

motorneuronal BMP, I evaluated Gbb processing in S2 cells and in cmpy mutant larvae by immunoprecipitation. I observed no differences in Gbb processing in the presence or absence of Cmpy, respectively. Nor did I find a decrease in BMP activity secreted into conditioned-media utilizing a well-established in vitro BMP signaling assay (Serpe et al. 2008). Together these data suggest that Cmpy does not distinguish motorneuronal Gbb by altering ligand processing and changing its appearance at the NMJ. Moreover, Cmpy does not inhibit Gbb secretion from S2 cells.

To further my understanding of Cmpy action based upon its localization, I generated a Cmpy-Venus transgene and evaluated its expression in larval motorneurons. Cmpy-Venus localizes within motorneuron cell bodies of the CNS and at the NMJ, where it is restricted to the presynaptic compartment. Cmpy-

Venus strongly colocalizes with a Gbb-HA transgene at the NMJ and in discrete

puncta within motor nerves, suggesting that Cmpy is unlikely to prevent Gbb

trafficking to the periphery. Surprisingly, Cmpy was necessary and sufficient to

promote Gbb-HA localization at the NMJ.

While Cmpy does not appear to affect BMP processing or prevent BMP

trafficking to the NMJ, the possibility remained that Cmpy regulates the timing of

motorneuronal BMP secretion to distinguish it from muscle-derived ligand. As

mentioned, BDNF and TGFβ signaling both regulate synaptic transmission and

both ligands undergo activity-dependent secretion (Thomas and Davies 2005,

Lacmann et al. 2007, Krieglstein et al. 2011). Presynaptic BDNF is sorted into

DCVs of the RSP in vertebrate neurons by the sorting receptors, Sortilin and

Carboxypeptidase E (Chen et al. 2005, Lou et al. 2005). However sorting receptors that direct TGFβs into the RSP for activity-dependent secretion have not been identified. Intriguingly, Sortilin interacts with the prodomain of BDNF, which is reminiscent of the specific interactions of Cmpy with precursor Gbb mentioned previously.

I hypothesized that Cmpy is a sorting receptor that directs presynaptic Gbb into the RSP in larval motorneurons. To test this hypothesis, I analyzed colocalization of Cmpy-Venus with a marker for DCVs of the RSP in Drosophila motorneurons.

In addition to strong colocalization between Cmpy-Venus and Gbb-HA, Cmpy-

Venus also strongly colocalizes with DCVs at the NMJ and in motor nerves, but

not with recycling neurotransmitter-containing synaptic vesicles at the NMJ.

Additionally, Gbb-HA signal at the NMJ is significantly diminished after high K+ depolarization, and this effect is dependent upon extracellular Ca2+. Gbb-HA levels are also diminished after driving synaptic activity with the blue light- activated cation channel, channelrhodopsin-2, consistent with activity-dependent

BMP secretion.

As previously mentioned neuronal Gbb is necessary for basal synaptic transmission at the NMJ (McCabe et al. 2003, Goold and Davis 2007). Many pathways that regulate synaptic function are conserved between vertebrates and invertebrates (Collins and DiAntonio 2007). Given that BDNF and TGFβ secretion from vertebrate neurons is regulated by activity and both modulate synaptic transmission, and Gbb levels at the NMJ are regulated by activity, I propose that activity-dependent, motorneuronal Gbb signaling similarly promotes neurotransmission at the NMJ. If Cmpy sorts motorneuronal Gbb into the RSP so that activity-dependent motorneuronal BMP signaling can modulate synaptic transmission, an obvious prediction is that synaptic function at cmpy mutant

NMJs should be impaired. Accordingly, I found that presynaptic neurotransmitter release is impaired at cmpy mutant NMJs, evidenced by a 28% decrease in evoked synaptic transmission and quantal content.

Future investigation into the interactions between presynaptic activity-dependent

BMP secretion and BMP pathway components available to transduce the activity- dependent cascade should begin to unravel the mechanisms by which neuronal

BMP is perceived as pro-function, as opposed to pro-growth, at the NMJ.

Further, Cmpy is the first identified TGFβ sorting receptor to date. Given the high degree of growth factor signaling pathway conservation between vertebrates and invertebrates, it is likely that developing a mechanistic understanding of Cmpy function at the tractable Drosophila NMJ will provide insight into roles of the mammalian homologs of Cmpy.

Figure 1.1

Dorsal

Ventral

Figure 1.1: Organization of the Larval Motor System

(A) Motorneuron (blue) and interneuron (pink) cell bodies, reside in the cortex

(grey) of the larval ventral ganglion. Motorneurons project into the central neuropil, where they branch to lay down dendrites and send an axon into the periphery to innervate their muscle targets. Adapted from (Tripodi and Arber

2012) with permission from Elsevier. (B) Motoraxons project ispi- (yellow, dark blue, tan, light/dark green, red) or contralaterally (light blue) out of the nerve cord in the transverse nerve (TN), intersegmental nerve (ISN), or segmental nerve

(SN) to innervate 30 muscles (grey) in each hemisegment of the body wall. A single hemisegment is shown, anterior is to the left; dorsal is at the top. Adapted from (Landgraf and Thor 2006) with permission from Elsevier.

Figure 1.2

Figure 1.2: Canonical BMP Signal Transduction

BMP ligand dimers (teal, Gbb at the Drosophila NMJ) bind to dimers of type I and type II serine-threonine kinase receptors (blue, Thickveins/Saxophone and

Wishful thinking, respectively), allowing phosphorylation (red circle) of the type I receptor by constitutively-active type II receptors. Phosphorylated, activated type

I receptors phosphorylate and active R-Smads (blue rectangle, Mad), promoting association with the Co-Smad (teal rectangle, Medea). This complex then translocates to the nucleus to bring about changes in gene transcription. Adapted from (Keshishian and Kim 2004) with permission from Elsevier.

Figure 1.3

Figure 1.3: BMP Signaling Pathways at the Drosophila NMJ

BMP ligand Gbb derived from the muscle (green) promotes growth (blue box) of

the motorneuron terminal through a canonical BMP signal transduction cascade.

Ligand binding to the type II BMP receptor Wit (blue) and type I receptors

Sax/Tkv (red) induces phosphorylation (yellow) of the R-Smad Mad (tan), its association with the Co-Smad Medea (grey), and translocation of this complex to the nucleus. In the nucleus these transcription factors associate with corepressors and/or co-activators to bring about changes in gene transcription that drive a general pro-growth program at the NMJ to add terminal synaptic boutons.

BMP ligand derived from motorneurons (purple) promotes synaptic transmission

(yellow box). The mechanisms by which motorneurons distinguish between the two sources of ligand are unclear.

CHAPTER 2:

Crimpy Inhibits the BMP Homolog Gbb in Motorneurons to Enable Proper

Growth Control at the Drosophila Neuromuscular Junction

(from James and Broihier, 2011)

Abstract

The BMP pathway is essential for scaling of the presynaptic motorneuron arbor to the postsynaptic muscle cell at the Drosophila neuromuscular junction (NMJ).

Genetic analyses indicate that the muscle is the BMP-sending cell and the motorneuron is the BMP-receiving cell. Nevertheless, it is unclear how this directionality is established since Glass bottom boat (Gbb), the known BMP ligand, is active in motorneurons. We demonstrate that crimpy (cmpy) limits neuronal Gbb activity to permit appropriate regulation of NMJ growth. cmpy was identified in a screen for motorneuron-expressed genes and codes for a single- pass transmembrane protein with sequence homology to vertebrate Crim1, or

Cysteine-rich in motorneurons1. We generated a targeted deletion of the cmpy locus and find that loss-of-function mutants exhibit excessive NMJ growth. In accordance with its expression profile, tissue-specific rescue experiments indicate that cmpy functions neuronally. The overgrowth in cmpy mutants is strictly dependent on activity of the BMP type II receptor Wishful thinking, arguing that Cmpy acts in the BMP pathway upstream of receptor activation and raising the possibility that it inhibits Gbb activity in motorneurons. Indeed, the cmpy mutant phenotype is strongly suppressed by RNAi-mediated knockdown of Gbb in motorneurons. Furthermore, Cmpy physically interacts with the Gbb precursor protein—arguing that Cmpy binds Gbb prior to ligand secretion. Together, these studies demonstrate that Cmpy restrains Gbb activity in motorneurons. We present a model where this inhibition permits the muscle-derived Gbb pool to

predominate at the NMJ, thus establishing retrograde directionality of the pro-

growth BMP pathway.

Introduction

Motorneurons constitute a fundamental line of communication between the CNS

and the periphery. In an anterograde direction, they integrate central interneuron

inputs to appropriately depolarize postsynaptic muscle to trigger contractions and

stimulate movement. In the retrograde direction, they translate information about

muscle activity to modulate synaptic size and strength at the neuromuscular

junction (NMJ). Thus, the NMJ is not only a location of neurotransmitter release,

but also a primary site of action for pathways that foster communication between

synaptic partners. Whereas the directionality of neurotransmission is defined by

the inherent cellular asymmetry of pre- and postsynaptic compartments, the

directionality of signaling pathway action at the synapse cannot be established in

the absence of functional analyses of individual pathway components.

The Wingless/Wnt and bone morphogenetic protein (BMP) morphogens mediate

coordinated differentiation of the motorneuron and the muscle cell at the

Drosophila NMJ (Packard et al. 2002, McCabe et al. 2003). During larval development new synaptic boutons are added to the motorneuron terminal in response to an increased demand for synaptic input as muscle area increases

(Atwood et al. 1993, Schuster et al. 1996, Schuster et al. 1996). Forward and reverse genetic approaches have defined pathways that regulate the

developmental expansion of the NMJ. Wingless is released from motorneuron terminals and binds to dFrizzled2 receptors on both the pre- and postsynaptic sides to stimulate NMJ growth and differentiation (Packard et al. 2002, Ataman et al. 2008, Miech et al. 2008).

On the other hand, the BMP homolog Glass bottom boat (Gbb) has been proposed to act in a retrograde manner to regulate synaptic growth and function

(McCabe et al. 2003). Gbb is postulated to be secreted from the muscle and to bind a type II BMP serine/threonine kinase receptor, Wishful thinking (Wit) on presynaptic motorneuron terminals. The Gbb-Wit interaction drives recruitment and activation of a type I receptor, either Saxophone (Sax) and/or Thickveins

(Tkv) (Aberle et al. 2002, Marques et al. 2002, Allan et al. 2003, McCabe et al.

2003, Rawson et al. 2003, McCabe et al. 2004). Signal transduction within the motorneuron acts via phosphorylation of the R-Smad, Mothers against

Decapentaplegic (Mad), the association of phospho-Mad with the co-Smad,

Medea (Med), and translocation of this complex to the nucleus to elicit changes in gene transcription (Raftery and Sutherland 1999, Keshishian and Kim 2004).

Loss-of-function (LOF) mutants in BMP pathway components result in NMJ undergrowth and impaired basal synaptic transmission at the NMJ (Aberle et al.

2002, Marques et al. 2002, McCabe et al. 2003, Rawson et al. 2003, McCabe et al. 2004). Conversely, elevated BMP signaling found in LOF mutants for the inhibitory Smad, Daughters against decapentaplegic (Dad) or in larvae

expressing the constitutively-active type I receptor Tkv, results in substantial

expansion of the NMJ (Sweeney and Davis 2002, O'Connor-Giles et al. 2008).

Furthermore, identification of factors that modulate BMP signaling activity on the

presynaptic side demonstrates that growth of the motorneuron arbor is

exquisitely sensitive to neuronal levels of BMP signal transduction (Sweeney and

Davis 2002, Wang et al. 2007, O'Connor-Giles et al. 2008, Kim et al. 2010).

Modes of BMP regulation at the NMJ extend to the postsynaptic compartment,

where the adaptor protein dCIP4 attenuates growth by inhibiting Gbb secretion

via a Wasp-Arp2/3 dependent mechanism (Nahm et al. 2010). Additionally, the

BMP pathway may serve an anterograde or autocrine function in muscle as Tkv and phospho-Mad are present in the post-synaptic compartment (Dudu et al.

2006). However, a function has not been assigned to this pathway, as presynaptic, but not postsynaptic expression of Mad, Med, Tkv, Sax and Wit, rescue the anatomical NMJ defects in the corresponding LOF mutants (Aberle et al. 2002, McCabe et al. 2004). Hence, components of the BMP signal transduction cascade are required in motorneurons for developmental NMJ expansion.

While a number of lines of evidence indicate that motorneurons receive a BMP signal, the source of the signal is less well established. The BMP homolog Gbb is postulated to act retrogradely on the basis of tissue-specific rescue experiments demonstrating that muscle-specific, but not neuronal-specific, expression of Gbb

in a hypomorphic gbb background rescues NMJ size and bouton number

(McCabe et al. 2003). However, neurotransmitter release at the NMJ is not rescued strongly in these animals (McCabe et al. 2003, Goold and Davis 2007).

In contrast, basal neurotransmission is fully recovered when Gbb is expressed pan-neuronally in a gbb deficient background (McCabe et al. 2003, Goold and

Davis 2007), suggesting the possibility of a presynaptic function for Gbb at the

NMJ. Consistent with this model, Gbb is expressed ubiquitously in late embryos

(McCabe et al. 2003). Moreover, a motorneuronal function for Gbb in larvae is strongly implied by functional studies demonstrating that Gbb acts retrogradely in motorneurons to strengthen synaptic transmission with their presynaptic partners.

This elegant work established that motorneuronal Gbb is necessary and sufficient to facilitate synaptic excitation between larval motorneurons and presynaptic cholinergic interneurons (Baines 2004).

We identified CG13253, which we named crimpy (cmpy), in a screen for

embryonic motorneuron-expressed transcripts. cmpy is predicted to encode a

cysteine-rich repeat (CRR) containing single-pass transmembrane protein, with

sequence homology to vertebrate Crim-1 (Cysteine-rich in motorneurons-1)

(Kolle et al. 2000, Kolle et al. 2003, Wilkinson et al. 2003). CRRs are present in a

large number of BMP-interacting proteins in vertebrates and invertebrates

(Umulis et al. 2009, Walsh et al. 2010). This structurally related family includes

extracellular antagonists such as Drosophila Short gastrulation (Sog) and

vertebrate Chordin (Chrd), believed to interfere with receptor-ligand interactions

(Francois et al. 1994, Bachiller et al. 2000). It also includes proteins such as

Gremlin and Sclerostin that can interact with BMPs intracellularly and are thought to interfere with BMP activity, at least in part, by altering ligand activation or secretion (Sun et al. 2006, Krause et al. 2010). We generated a targeted deletion

of the CG13253/cmpy locus and find that homozygotes exhibit aberrant NMJ

differentiation with increased numbers of type I boutons. Here we present

evidence that Cmpy is a novel antagonist of BMP signaling at the NMJ. We

propose that Cmpy antagonizes motorneuronal Gbb activity to establish

retrograde directionality of the pro-growth Gbb signal, hence maintaining

synchronization of presynaptic axon elaboration and postsynaptic muscle growth.

Material and Methods

Fly Stocks

Stocks used in this work include: UAS-CG13253RNAi-KK library (transformant ID

101249) Vienna Drosophila RNAi center, ElavGal4, D42Gal4 and OK6Gal4 (A.

DiAntonio), Mad1 and A9Gal4, (K. O’Connor-Giles), UAS-gbb(9.9) (B. McCabe),

UAS-gbbRNAi (K. Wharton), UAS-cmpy transgenic flies were generated by

BestGene, Inc. All other stocks were obtained from Bloomington Stock Center.

Identification of CG13253 and allele generation

To identify uncharacterized genes with neuronal expression, an in silico screen of

the FlyExpress expression pattern database was performed. We were interested

in transcripts with embryonic nerve cord expression patterns that were not pan-

neuronal, with the goal of identifying genes acting in functionally related neuronal

populations. We obtained ESTs from the Drosophila Genomics Resource Center

for genes with apparent expression in specific neuronal subtypes and analyzed

their RNA expression patterns. Among this group, CG13253 was of interest since

its spatiotemporal expression pattern at stage 14 resembled that of Hb9 (Broihier

and Skeath 2002).

To generate a null allele of CG13253, targeted recombination between FRT-site containing piggyBac elements was utilized (Parks et al. 2004, Thibault et al.

2004). In the presence of FLP recombinase, recombination between the FRT

sites of heterozygous piggyBac elements in trans in germline stem cells can lead

to recovery of progeny carrying deletions for the genomic region flanked by the

two piggyBacs. The resulting progeny will contain a hybrid piggyBac element

containing portions of each parental P-element. In this manner, an approximately eight kb region was deleted between PBac{WH}f02482 and PBac{WH}f01736 at

77E3 on chromosome 3L (CG13253∆8; Fig. 2.1F), including the translation initiation site and most of the protein coding region of CG13253. Four approaches were utilized to verify the CG13253∆8 deletion. (1) PCR to amplify from each side of the resulting hybrid PBac into genomic DNA verified the presence of portions of both elements within the same genetic background, indicating a recombination event (data not shown). (2) Genomic primers targeting the putatively deleted region were used to amplify an approximately three kb band from Oregon R (OR) DNA by PCR which is absent from CG13253∆8

homozygous DNA (data not shown). PCR with genomic primers flanking the resultant hybrid PBac verify deletion by demonstrating the size difference in intervening DNA in OR and CG13253∆8 homozygous DNA (Fig. 2.1G). (4) in situ hybridization analysis of CG13253 homozygous mutant embryos reveals an absence of CG13253 mRNA (Fig. 2.1H).

A second gene, CG34260, is annotated within the CG13253 locus on FlyBase and is present within the eight kb region. CG34260 is located on the opposite

DNA strand as CG13253 and is predicted to code for a 219 amino acid protein.

However, no ESTs have been identified and there are no known structural domains. We generated an antisense RNA probe to CG34260 and do not detect embryonic expression (data not shown). These data argue that CG34260 is unlikely to be an embryonically expressed transcript. Since CG13253 RNAi phenocopies the NMJ phenotype displayed by CG13253∆8 homozygotes, and a

CG13253 transgene rescues the NMJ phenotype in deletion animals, we conclude that the NMJ phenotypes present in deletion animals result from loss of

CG13253 function.

Immunohistochemistry

Embryo fixation, in situ hybridization, and immunohistochemistry were carried out as previously described (Miller et al. 2008). The Roche DIG RNA Labeling Kit

(SP6/T7) was utilized to make digoxigenin (DIG)-labeled RNA probe using full length CG13253 cDNA as template. For all larval experiments, 10 virgin females

were crossed to 5 males, and bottles were maintained at 25°C for four days

before removing adults. Dissection of wandering third instar larvae was carried

out in ice-cold PBS, and body walls were fixed in Bouin’s Fixative (Polysciences

Inc.). The following primary antibodies were used: rabbit anti-Hb9 at 1:1000,

rabbit anti-Eve at 1:1000, mAb 1D4 at 1:10 (anti-Fasciclin II; Developmental

Studies Hybridoma Bank), rabbit anti-HRP at 1:300 (Jackson Laboratories), mouse anti-Dlg at 1:1000, mouse anti-NC82 (Bruchpilot) at 1:100, rabbit anti-

Nwk at 1:1000 (K. O’Connor-Giles), and rabbit anti-dGluRIII at 1:5000 (A.

DiAntonio). Biotinylated secondary antibodies were used at 1:300 for immunohistochemistry, and Alexa Fluor 488 and 568 (Invitrogen) were used at

1:300 for immunofluorescence.

Imaging and Data Analysis

Embryos and larvae were dissected using a Leica MZ9 dissecting microscope and analyzed on a Zeiss Axioplan 2 microscope with 40X, 63X, and/or 100X oil- immersion objectives. Images of embryonic ventral nerve cords were captured with an AxioCam MRc camera, and fluorescence images of larval body walls were obtained with a Zeiss Axio Imager.ZI confocal microscope at 40X.

Brightness and contrast were adjusted in Adobe Photoshop CS5. Quantification of type I glutamatergic boutons was carried out at NMJ 6/7 and NMJ 4. The muscle area of all genotypes analyzed was similar. For NMJ 6/7, boutons were quantified at both segment A2 and segment A3. Although consistent results were obtained for all experiments at A3, only data for NMJ 6/7 in A2 are presented

given segment-specific differences. The data presented for NMJ 4 are pooled

from both A2 and A3. Satellite boutons at NMJ 4 were defined as either

extensions of three or fewer boutons projecting from the primary branch of the

nerve terminal, or as single boutons that bud off of primary boutons without an

intervening axon segment. Groups of means were compared via one-way

ANOVA, and the unpaired Student’s t-test was used for comparisons between pairs of means. *, p<0.05; **, p<0.01; p<0.001. n.s., not significant.

Plasmids

EST clones for cmpy (RE53920), gbb (GH12092), pAWF C-terminal 3X Flag tag, and pAWH C-terminal 3X HA tag, Gateway System (Invitrogen) compatible vectors were obtained from the Drosophila Genomics Resource Center. Vectors pDEST-GBKT7 and pDEST-GADT7 (Rossignol et al. 2007) for yeast-two hybrid were obtained from ABRC DNA stock center. For immunoprecipitation experiments, full-length cmpy and gbb coding sequences in the pCR8/GW/TOPO vector (Invitrogen) were cloned into pAWF or pAWH destination vectors with

Gateway Long Range Clonase II Enzyme Mix according to manufacturer’s protocols (Invitrogen). For yeast two-hybrid experiments, domains were subcloned into pDEST-GADT7 (cmpy constructs, prey) or pDEST-GBKT7 (gbb constructs, bait). Immunoprecipitation and yeast two-hybrid experiments were carried out as previously described (Weng et al. 2011).

Results

CG13253 encodes a predicted single-pass transmembrane protein

expressed in the CNS

CG13253 was identified in a screen for embryonic motorneuron-expressed transcripts in the ventral nerve cord (VNC) (Materials and Methods). CG13253

RNA expression initiates at stage 13 in clusters of medial and lateral post-mitotic

neurons in the VNC. At embryonic stage 14, it is expressed in a segmentally

repeated V-shaped pattern (Fig. 2.1A) reminiscent of the expression profile of

Hb9, a marker for ventrally and laterally projecting motorneurons (Fig. 2.1B)

(Broihier and Skeath 2002, Odden et al. 2002). Indeed, CG13253 is co-

expressed with Hb9 in a subset of Hb9-positive neurons, including the ventrally

projecting RP motorneurons (Fig. 2.1C), as well as dorsally projecting Even-

skipped positive motorneurons (data not shown) (Landgraf et al. 1999). The

CG13253 expression domain expands at stage 15, and by the end of

embryogenesis it is widely expressed in the VNC (Fig. 2.1D). In the embryo,

appreciable CG13253 expression is not detected outside of the VNC (data not

shown). Hence, CG13253 is expressed in dorsally and ventrally motorneuron

populations, though the large number of CG13253-positive cells in the VNC

indicates that its expression is not motorneuron-specific.

CG13253 is predicted to encode a single-pass 273 amino acid type II

transmembrane protein with an insulin-like growth factor binding protein (IGFBP)- like domain and a single low-threshold cysteine-rich repeat (CRR; Fig. 2.1E). We

also noticed an arginine-lysine rich domain at the C terminus. CG13253 shares

homology with vertebrate Crim-1, or Cysteine-rich in motorneurons-1 (Kolle et al.

2000, Kolle et al. 2003). Vertebrate Crim-1 codes for a single-pass transmembrane protein with an N-terminal IGFBP-like domain, a short C-terminal cytoplasmic tail, and six CRRs interspersed between the IGFBP motif and the putative transmembrane domain (Kolle et al. 2000). Crim-1 is expressed in early populations of motorneurons and interneurons in the developing mouse, though its neuronal function remains obscure (Wilkinson et al. 2003).

CG13253, or crimpy, functions in motorneurons to restrict NMJ growth

To investigate CG13253 function in neuronal development, a loss-of-function

(LOF) allele was generated by targeted recombination between piggyBac elements in trans (Parks et al. 2004). In this manner we generated an 8 kb deletion (CG13253∆8), including roughly 3 kb upstream, the translation initiation site, and the majority of coding sequence (Fig. 2.1F). Deletion was verified by multiple PCR-based strategies (Materials and Methods), including amplification across the deletion with genomic primers to either side of the resulting hybrid

PBac element. Amplification in a wild-type, non piggyBac-containing background yields an 8.89 kb product, while amplification across the hybrid piggyBac that remains following recombination yields a 7.86 kb product. Of this, 7.23 kb corresponds to the hybrid piggyBac, and 630 bases correspond to genomic DNA

(Fig. 2.1G). Consistent with the presence of the deletion, CG13253 RNA is not expressed in homozygous deletion embryos (Fig. 2.1H), indicating that this

deletion represents a null allele. CG13253∆8 mutants are viable, although the homozygous females are sterile. In accordance with characteristics of the mutant phenotype described below, and sequence similarity of CG13253 to vertebrate

∆ Crim-1, we name this gene crimpy (cmpy), and refer to the LOF allele as cmpy 8.

Since cmpy is expressed in motorneurons when cell fate and motor axon guidance decisions occur, we tested whether cmpy regulates these processes.

We do not observe defects in neuronal cell fate acquisition or axon guidance in cmpy homozygotes, as assessed by cell fate markers Even-skipped and Hb9, and axonal marker mAb 1D4 (data not shown) (Vactor et al. 1993, Landgraf et al.

1999, Broihier and Skeath 2002). These results suggest that cmpy does not contribute to motorneuron cell fate specification or axon guidance, motivating us to examine later stages of motorneuron differentiation.

The fly NMJ serves as an ideal model for the investigation of synaptic development and function (Collins and DiAntonio 2007). Since cmpy is expressed in motorneurons, we asked whether cmpy mutants exhibit defects in

NMJ development. We scored all type I glutamatergic boutons (Johansen et al.

1989) at two well-characterized identifiable NMJs—the NMJ that innervates the cleft between muscles 6 and 7 (NMJ 6/7), and the NMJ on the face of muscle 4

(NMJ 4). cmpy∆8 homozygotes display a 52% increase in the number of boutons at NMJ 6/7 and a 57% increase in type I boutons at NMJ 4 (Fig. 2.2A-D,I,J, Table

2.1), indicating that cmpy restrains NMJ growth. We observe comparably

increased bouton number in cmpy∆8/Df(3L)452 larvae (Fig. 2.2I,J; Table 2.1),

supporting the conclusion that cmpy∆8 is a null allele. Type I boutons are further

classified into two types based on their size and the extent of the Discs-large

(Dlg)-positive post-synaptic sub-synaptic reticulum (SSR) (Atwood et al. 1993,

Lahey et al. 1994). Type Ib (big) boutons are surrounded by a prominent Dlg- positive SSR, whereas the Dlg-positive SSR enveloping type Is (small) boutons is less extensive. Although cmpy mutant boutons tend to be smaller than those in wild type, the overall proportion with strong Dlg immunofluorescence appears unchanged (data not shown), arguing that cmpy does not selectively regulate the development of type Ib or type Is boutons. We further quantified the number of satellite boutons at NMJ 4 in cmpy homozygotes (Dickman et al. 2006,

O'Connor-Giles et al. 2008). We define satellite boutons as small boutons present on short branches (three or fewer boutons) distinct from primary arbors. cmpy mutants display a 2.9-fold increase in the number of satellite boutons at

NMJ 4 (Fig. 2.2K). A comparable increase in satellite bouton formation is

observed in mutants with elevated levels of BMP signaling (O'Connor-Giles et al.

2008). The presence of numerous small boutons in cmpy mutants is reminiscent

of a type of rock climbing route featuring small holds, or crimps—which can be

described as a “crimpy” route.

Since cmpy codes for a neuronal transcript, a straightforward hypothesis is that

cmpy acts in motorneurons to inhibit NMJ growth. Thus, we utilized the

GAL4/UAS transactivation system (Brand and Perrimon 1993) to evaluate the

ability of RNAi-mediated knockdown of cmpy to recapitulate the cmpy∆8

overgrowth phenotype. Using the larval motorneuron driver D42Gal4 (Sanyal

2009), we find 47% and 49% increases in bouton number at NMJ 6/7 and NMJ 4,

respectively, in D42>cmpyRNAi larvae compared to controls (Fig. 2.2A-B, E-F, I-J;

Table 2.1). We observe comparable increases in bouton number using the pan-

neuronal driver ElavGal4 to drive cmpy RNA knockdown (data not shown). The

NMJ overgrowth displayed by animals with neuronal-specific cmpy knockdown argues that cmpy acts presynaptically. As a key test of this hypothesis, we performed tissue-specific rescue experiments. Since neuronal cmpy overexpression in an otherwise wild-type background does not alter NMJ growth

(Table 2.2), we examined whether D42Gal4-mediated overexpression of cmpy in a cmpy homozygous background rescues proper regulation of NMJ growth. At

NMJ 4, proper bouton number is fully rescued, whereas at NMJ 6/7 overgrowth is reduced from 52% in cmpy homozygotes to 9% in cmpy homozygotes with motorneuronal cmpy expression (Fig. 2.2G-J; Table 2.1). Comparable rescue was observed using ElavGal4 to drive cmpy in all neurons (Table 2.1). Thus, cmpy acts in motorneurons to limit NMJ arbor expansion during development.

In addition to characterizing NMJ growth, we analyzed synaptic structure in cmpy∆8 larvae. To evaluate active zone formation, we used the antibody against

Bruchpilot (Wagh et al. 2006) to label active zones. Indeed, cmpy mutant NMJs contain more active zones. However, the percent increase is comparable to the percent increase in bouton number so that active zone density remains constant

(data not shown). Furthermore, assembly of periactive zones, evidenced by

localization of the endocytic adaptor Nervous wreck, appears normal in cmpy

mutants (data not shown). On the postsynaptic side, an antibody to the glutamate

receptor subunit DGluRIII (Marrus et al. 2004) was utilized to mark glutamate receptor clusters. We do not detect a change in glutamate receptor distribution in cmpy mutants (data not shown). Hence, cmpy does not appear to function selectively in synapse assembly or maturation, but rather appears necessary for regulation of a general NMJ growth-promoting program.

cmpy is genetically upstream of the BMP Type II receptor wit

The BMP pathway has emerged as a critical positive regulator of NMJ growth.

LOF mutants in multiple pathway components, including gbb, wit, tkv, sax, mad,

and medea all exhibit small NMJs with reduced numbers of boutons (Aberle et al.

2002, Marques et al. 2002, Packard et al. 2002, Rawson et al. 2003, McCabe et

al. 2004). On the other hand, mutants with elevated BMP activity, such as

mutants for the inhibitory Smad, Dad, display NMJ overgrowth, arguing that

levels of BMP activity instruct arbor expansion (Sweeney and Davis 2002,

O'Connor-Giles et al. 2008).

To evaluate whether the increased number of boutons present in cmpy mutants

reflects BMP pathway dysregulation, we tested if the cmpy LOF mutant

phenotype is dominantly suppressed by LOF mutations in genes with pro-BMP

activity. The BMP ligand gbb, the type II receptor wit and the transcription factor

Mad are essential for BMP signaling at the NMJ, and NMJs in corresponding

LOF homozygotes are undergrown (Aberle et al. 2002, Marques et al. 2002,

McCabe et al. 2003, Rawson et al. 2003). We do not observe defects in NMJ

expansion in gbb, wit, or Mad heterozygotes (Table 2.1). Yet the overgrowth

observed in cmpy mutants is suppressed by loss of one wild-type copy of gbb,

wit or Mad (Fig. 2.3A-H, K, L; Table 2.1). For example, at NMJ 6/7 the percent

increase in bouton number falls from 52% in cmpy homozygotes to 24% in cmpy

mutants with loss of one copy of either wit or Mad, and to 20% with loss of one copy of gbb. These dominant genetic interactions argue that cmpy attenuates

NMJ expansion by inhibiting BMP pathway activity.

While these experiments place cmpy in the BMP pathway, they do not address the order of gene action. Thus, we conducted a classic genetic epistasis experiment to determine whether the increase in bouton number displayed by cmpy LOF mutants depends on activity of the BMP type II receptor wit. We find that wit cmpy double mutants phenocopy the NMJ undergrowth displayed by wit

single mutants (Fig. 2.3I-L; Table 2.1). At NMJ 6/7, we find a 38% reduction in

bouton number in wit mutants relative to a 33% reduction in wit cmpy doubles.

Likewise at NMJ 4, wit mutants exhibit a 46% reduction in bouton number compared to 44% in wit cmpy doubles. Hence, wit mutants fully suppress the

NMJ overgrowth observed in cmpy mutants, placing cmpy genetically upstream of wit in a common signaling pathway. Since cmpy regulates NMJ growth

upstream of BMP receptor activity, we probed the relationship between cmpy and

the BMP ligand gbb.

cmpy overexpression blunts the phenotypic consequences of gbb overexpression

Muscle-specific expression of Gbb rescues the reduction in bouton number

exhibited by gbb LOF mutants, demonstrating that the pathway can act in a

retrograde direction (McCabe et al. 2003). However, Gbb expression is not

muscle-specific (McCabe et al. 2003). In particular, Gbb functions in the VNC

(Baines 2004). Motorneuronal Gbb poses a potential dilemma for models of

retrograde BMP signaling at the NMJ. If Gbb constitutes the critical extracellular

cue informing the presynaptic motorneuron about the size of its postsynaptic

muscle partner, then it would appear critical for the motorneuron to perceive

primarily the muscle-derived Gbb pool to properly regulate growth. If Gbb were

secreted from the motorneuron terminal, it would effectively dilute the muscle-

derived pool, and potentially decouple pre- and postsynaptic growth. Hence, we

reasoned that motorneurons might possess a cellular mechanism to modulate or

inhibit Gbb at the NMJ.

As a test of this hypothesis, we overexpressed Gbb with the motorneuron driver

D42Gal4 and assessed NMJ morphology. D42Gal4/UAS-gbb larvae exhibit 40%

and 62% increases in bouton number at NMJs 6/7 and 4, respectively (Fig. 2.4A-

D, G, H; Table 2.2). Bouton numbers were comparably increased when Gbb was

overexpressed via the OK6Gal4 motorneuron driver (data not shown) (Sanyal

2009). This overgrowth demonstrates that excess Gbb in motorneurons

overwhelms growth regulatory mechanisms at the NMJ. The percent increase in

bouton number observed with motorneuronal gbb overexpression is similar to

that displayed by cmpy LOF mutants, and genetic epistasis experiments place

cmpy in the BMP pathway upstream of the wit receptor—arguing that cmpy may

inhibit gbb. Thus, we examined whether cmpy overexpression suppresses gbb-

dependent NMJ overgrowth. Remarkably, co-overexpression of cmpy and gbb in

motorneurons with either D42Gal4 or OK6Gal4 drivers results in NMJs with wild- type numbers of type I boutons at both NMJ 6/7 and NMJ 4 (Fig. 2.4E-H, Table

2.2; data not shown). The Gbb overexpression phenotype is not suppressed by co-overexpression of lacZ, indicating that suppression of Gbb-induced overgrowth is mediated by Cmpy (Fig. 2.4G-H; Table 2.2). Importantly, neuronal cmpy overexpression does not inhibit NMJ growth in an otherwise wild-type background (Table 2.2), indicating that the suppression reflects an intimate relationship between Gbb and Cmpy and is not a secondary consequence of generic Cmpy-dependent growth inhibition. We conclude that neuronal Gbb overexpression likely drives excessive NMJ growth by an autocrine mechanism.

Cmpy-dependent suppression of this phenotype argues that Cmpy inhibits the growth-promoting activity of Gbb in motorneurons.

To investigate whether the Gbb-inhibiting activity of Cmpy is motorneuron- specific, we investigated the ability of cmpy to antagonize gbb activity in an

independent cellular context. Overexpression of Gbb in the developing wing

imaginal disc results in a wing blistering phenotype (Khalsa et al. 1998). We find

that Gbb overexpression in the wing disc with A9Gal4 results in a blistering

phenotype in 79% of females, and a more severe phenotype, including unfurling

and blistering of the wing, in 93% of males (Fig. 2.5A-D; Table 2.3). Co-

misexpression of cmpy and gbb strongly suppresses the blistering phenotype.

Only 11% of A9Gal4/UAS-gbb, UAS-cmpy females had blistered wings, whereas

in males the severity of the unfurling/blistering phenotype is lessened, such that

89% of males exhibit wing blistering, and only 11% exhibit unfurling of the wing

and blistering (Fig. 2.5E-F; Table 2.3). The suppression is specific, as wing

phenotypes are not suppressed by co-overexpression of lacZ (Table 2.3). Hence,

cmpy overexpression suppresses gbb-dependent phenotypes in two cellular

contexts, larval motorneurons and the wing imaginal disc.

Cmpy antagonizes motorneuron-derived Gbb at the NMJ

Our genetic analyses suggest that cmpy antagonizes BMP pathway activity in

motorneurons. If so, the NMJ overgrowth observed in cmpy LOF mutants should

be suppressed by RNAi-mediated knockdown of Gbb in motorneurons. We first

tested whether motorneuronal Gbb is necessary for normal NMJ growth by

driving UAS-gbbRNAi (Ballard et al. 2010) in motorneurons using the D42Gal4

driver. D42Gal4/UAS-gbbRNAi larvae display wild-type numbers of boutons at NMJ

6/7 and NMJ 4 (Fig. 2.6A-D, K-L; Table 2.4), suggesting that motorneuron- derived Gbb does not regulate bouton number. Supporting this conclusion,

neuronal cmpy overexpression driven with either ElavGal4 or D42Gal4 has no effect on bouton number (Table 2.2; data not shown). These data argue that presynaptic Gbb does not play a role in NMJ growth regulation. To investigate efficacy of the gbbRNAi construct, we also expressed it postsynaptically.

24BGal4/UAS-gbbRNAi animals display 28% decreases in bouton number at both

NMJ 6/7 and NMJ 4 (Fig. 2.6E-F, K-L; Table 2.4), providing evidence that UAS-

gbbRNAi inhibits Gbb expression. This experiment also establishes that muscle-

derived Gbb is necessary for proper regulation of NMJ expansion.

We next assessed whether motorneuronal expression of gbbRNAi modulates the

NMJ overgrowth in cmpy LOF mutants. Driving gbbRNAi using D42Gal4 in a cmpy

homozygous mutant background strongly suppresses the cmpy LOF mutant

phenotype. RNAi-mediated Gbb knockdown of Gbb in motorneurons in cmpy

mutants suppresses the NMJ overgrowth from 46% to 16% at NMJ 6/7 and from

49% to 20% at NMJ 4 (Fig. 2.6G-L; Table 2.4). These results indicate that Cmpy

inhibits the motorneuron-derived pool of Gbb at the NMJ.

Since Gbb is normally secreted from the postsynaptic muscle, muscle-specific

Cmpy misexpression is predicted to interfere with NMJ growth. Utilizing 24BGal4

to drive cmpy misexpression in muscle, we observe a 20% decrease in type I

boutons at NMJ 6/7 and a 22% decrease at NMJ 4 (Table 2.2). Comparable

results were obtained with a second UAS-cmpy line (data not shown). Thus,

muscle-specific Cmpy misexpression drives NMJ undergrowth, consistent with

the model that Cmpy antagonizes Gbb.

Cmpy physically interacts with the Gbb precursor protein

BMP activity is regulated at multiple levels, including processing, secretion, and

receptor binding (Umulis et al. 2009, Walsh et al. 2010). BMPs are synthesized

as precursor proteins that are cleaved by endopeptidases into prodomain and

signaling fragments (Cui et al. 1998, Constam and Robertson 1999, Kunnapuu et

al. 2009). After processing, the prodomain remains non-covalently associated

with the mature signaling fragment and may serve a regulatory function (Cui et al.

1998, Degnin et al. 2004). Additionally, CRR-containing BMP antagonists,

including vertebrate Chordin and Noggin, and Drosophila Short gastrulation (Sog)

predominantly inhibit BMP activity by binding to BMPs and blocking ligand-

receptor interactions (Umulis et al. 2009, Walsh et al. 2010).

To test whether Cmpy and Gbb associate in a complex, we performed co-

immunoprecipitation experiments from S2 cell lysates. Indeed, C-terminal epitope tagged Cmpy-Flag and Gbb-HA proteins co-immunoprecipitate (Fig. 2.7A). Cmpy immunoprecipitates unprocessed Gbb (55 kDa), indicating that Cmpy associates with full-length Gbb precursor protein. Gbb also immunoprecipitates Cmpy (Fig.

2.7A). We detect two forms of Cmpy in these experiments: a 33 kDa form, consistent with full-length protein, and a smaller 25 kDa form (Fig. 2.7A). The smaller Cmpy isoform suggests that Cmpy is proteolytically processed, while its

molecular weight suggests a cleavage site immediately C-terminal to the

predicted transmembrane domain (arrow in Fig. 2.7B). Gbb selectively

associates with the smaller 25 kDa Cmpy isoform, arguing that Gbb interacts

preferentially with processed Cmpy. Thus, Cmpy and Gbb can associate in a

complex. Furthermore, since Cmpy associates with the precursor protein form of

Gbb, these experiments suggest that Cmpy binds Gbb prior to generation of

mature ligand.

We turned to a yeast interaction assay to verify the relevant protein interaction

domains and assess the likelihood that the Cmpy-Gbb association is direct. We tested N- and C-terminal fragments of Cmpy and found that the C-terminal Cmpy fragment interacts with Gbb (Fig. 2.7B, C) consistent with the 25 kDa Cmpy isoform that co-immunoprecipitates with Gbb. Furthermore, the precursor form of

Gbb (Gbb-Pre) interacts robustly with Cmpy in yeast, while processed/mature

Gbb (Gbb-Mat) interacts only weakly (Fig. 2.7B, C). We next subdivided the

Cmpy-C fragment into a fragment containing the cysteine-rich repeat (Cmpy-CR),

and a fragment containing the lysine-arginine rich region (Cmpy-RK). Surprisingly,

the Cmpy-RK fragment interacts with Gbb-Pre, while Cmpy-CR interacts with neither Gbb-Pre nor Gbb-Mat (Fig. 2.7B, D), suggesting that the Cmpy-Gbb interaction is mediated by sequences outside of the CRR. These data support the co-immunoprecipitation results and indicate that Cmpy binds the unprocessed, precursor form of Gbb—in line with the model that Cmpy interacts with Gbb prior to ligand processing. Together, the experiments presented here

indicate that Cmpy inhibits Gbb in motorneurons, thus contributing to the

retrograde directionality of the pro-growth BMP signal at the NMJ.

Discussion

Gbb has been proposed to cue presynaptic motorneurons to the size of their postsynaptic muscle partners. However, muscles have not been established as the primary source of Gbb at the NMJ. In fact, motorneuron-derived Gbb has a

critical retrograde activity at the motorneuron-interneuron synapse (Baines 2004),

demonstrating that motorneuronal Gbb is active. In the present work, we

demonstrate that motorneurons express Cmpy, a Gbb antagonist. We propose

that Cmpy restrains motorneuronal activity of Gbb at the NMJ thus establishing

the muscle as the predominant source of the pro-growth BMP signal. Here we

discuss potential mechanisms for Cmpy function at the NMJ and Cmpy’s

relationship to intracellular and extracellular BMP antagonists.

Our interest in CG13253/Crimpy was sparked by its restricted expression in the

ventral nerve cord and was reinforced by the presence of a predicted

transmembrane domain and CRR. The presence of these two sequence

elements renders Cmpy similar to vertebrate Crim-1, which is a single-pass transmembrane domain protein with six CRRs (Kolle et al. 2000, Kolle et al.

2003). In mice, Crim-1 hypomorphs have been described and display pleiotropic defects in multiple organ systems (Pennisi et al. 2007). Notably, Crim-1 is expressed in developing motorneuron and interneuron populations in the

developing mouse and chick spinal cords, though LOF studies have not

addressed a neuronal function. A Crim-1 homolog has also been described in

zebrafish, where it is linked to vascular and somitic development (Kinna et al.

2006), and in C. elegans, where RNAi-mediated knockdown of crm-1 suggests a pro-BMP function in the control of body size (Fung et al. 2007). Cell culture studies provide evidence that Crim-1 binds BMP4/7 and antagonizes the production and processing of the preprotein in the Golgi (Wilkinson et al. 2003).

Interestingly, these authors also demonstrated that Crim-1 interacts with BMP4/7

at the cell surface and inhibits BMP secretion into the media (Wilkinson et al.

2003), raising the possibility that Crim-1 antagonizes BMP signaling by multiple

cellular mechanisms.

CRR containing proteins are established modulators of BMP signaling in

vertebrates and invertebrates. In Drosophila, posterior wing crossvein

specification requires local activation of the BMP pathway, and loss of BMP

signaling yields a crossveinless phenotype (Conley et al. 2000). BMP ligands are

produced in neighboring longitudinal wing veins and transported to the posterior

crossvein (Ray and Wharton 2001, Ralston and Blair 2005). Ligand activity is

differentially regulated by secreted CRR containing proteins Short gastrulation

(Sog) and Crossveinless-2 (Cv-2). Sog and Cv-2 both have pro- and anti-BMP activity, though their mode and range of action differ (Serpe et al. 2005, Shimmi

et al. 2005, Serpe et al. 2008). Sog is proposed to act at long range. Its anti-BMP

activity is thought to derive from sequestering BMPs from their receptors, while

its pro-BMP activity likely arises from transporting BMP ligands through tissues

(Serpe et al. 2005, Shimmi et al. 2005). In constrast, Cv-2 is proposed to act at short range and binds heparan sulfate proteoglycans and the type I receptor Tkv.

The biphasic activities of Sog and Cv-2 serve to emphasize the complex modes of extracellular regulation of BMPs by CRR containing proteins as well as to draw attention to possible differences between BMP regulation in the wing and Cmpy- dependent BMP regulation at the NMJ. While overexpression of Cmpy suppresses Gbb overexpression phenotypes in the wing, cmpy LOF mutants do

not display wing vein phenotypes (R.E.J. and H.T.B., unpublished data). Cmpy

also does not function during early embryogenesis when the BMP homolog

Decapentaplegic acts as a classical morphogen in dorsoventral patterning

(Francois et al. 1994, Decotto and Ferguson 2001, Mizutani et al. 2005). In both

the early embryo and the wing, BMP activity is shaped over many cell diameters

by extracellular CRR containing proteins. As discussed above, Sog and Cv-2

play essential extracellular roles in establishing the magnitude and directionality

of BMP signaling. In contrast, Gbb is proposed to act locally at the NMJ to couple

pre- and postsynaptic growth. In this model, the muscle is suggested to release

Gbb in proportion to its size, informing the presynaptic motorneuron about its

postsynaptic target (McCabe et al. 2003).

The close apposition of the BMP-sending and receiving cells at the NMJ may relieve a requirement for long-range extracellular regulation of the ligand. Instead, we propose that a primary challenge at the NMJ is to establish the cellular source

of the BMP signal since Gbb is present both in motorneurons and muscle

(Wharton et al. 1999, McCabe et al. 2003, Baines 2004). In this case, cell- autonomous regulation of the ligand could provide a mechanism for the motorneuron to discriminate between motorneuron- and muscle-derived pools.

Consistent with this model, we have presented evidence that Cmpy binds Gbb prior to processing and inhibits its growth-promoting activity in motorneurons. In this manner, the Cmpy-Gbb interaction might provide motorneurons with an effective mechanism for distinguishing autocrine and paracrine Gbb signals

within the NMJ microenvironment.

CRR containing BMP antagonists were initially identified for their extracellular

roles in the establishment of BMP morphogenetic gradients (Garcia Abreu et al.

2002, Zakin and De Robertis 2010). It will be interesting to determine if additional

CRR containing proteins function intracellularly as more short-range BMP-

dependent signaling interactions are thoroughly described. Consistent with this

idea, several mammalian CRR containing proteins bind precursor forms of BMP

and inhibit BMP activity or secretion in a cell-autonomous manner (Sun et al.

2006, Krause et al. 2010). Gremlin is a BMP antagonist expressed in

differentiated cells, including neurons (Topol et al. 1997). When co-expressed

with BMP4, it binds to the precursor form of BMP4 and inhibits secretion (Sun et

al. 2006). Sclerostin, another BMP antagonist, inhibits BMP7 secretion when the

proteins are co-expressed in osteocytes (Krause et al. 2010). These studies

argue that intracellular modulation of ligand production contributes to BMP signaling directionality in vertebrates.

The work presented here suggests that Cmpy antagonizes Gbb activity in motorneurons prior to ligand secretion. To further delineate the Cmpy-Gbb relationship, it will be important to map their localization patterns in motorneurons using compartment-specific markers. While attempts to generate anti-Cmpy antibodies have been unsuccessful (R.E.J. and H.T.B., unpublished data), generation of transgenic flies carrying epitope-tagged Cmpy may enable an analysis of Cmpy’s subcellular localization. Cmpy-mediated inhibition of Gbb at the NMJ may rely upon restricted localization of Cmpy to this subcellular locale; however, the possibility that Cmpy regulates Gbb’s activity at the central synapse remains open. Investigation of the localization pattern of Cmpy in motorneurons will begin to address the issue of Cmpy function at these distinct synapses.

Furthermore, an analysis of Gbb distribution, trafficking, and secretion in motorneurons in cmpy mutants will indicate the stage of Gbb processing at which

Cmpy is likely to act. Studies on mammalian Sclerostin provide precedent for an intracellular mechanism for BMP inhibition, as Sclerostin sequesters BMP7 preprotein, leading to its intracellular retention and proteasomal degradation

(Krause et al. 2010). Interestingly, Cmpy contains only a single low-threshold

CRR. These motifs modulate interactions with mature secreted ligand (Walsh et al. 2010), suggesting that sequences outside of the CRR mediate interactions

with the precursor form of Gbb. Indeed, Cmpy’s interaction with Gbb is dependent on C-terminal sequences including a lysine/arginine-rich domain at the extreme C terminus. Likewise, Gremlin’s intracellular interaction with the precursor form of BMP4 is not modulated by its cysteine-rich region, but rather by a lysine/arginine-rich domain (Sun et al. 2006). The sequence similarities between the BMP interaction domains in Gremlin and Crimpy raise the possibility that these proteins antagonize BMP activity by a conserved mechanism.

We have focused here on Cmpy regulation of Gbb in anatomical development of the NMJ. In addition, Gbb regulates baseline neurotransmission and synaptic homeostasis at the NMJ (Goold and Davis 2007). Motorneurons precisely compensate for impaired postsynaptic neurotransmitter receptor sensitivity by increasing presynaptic neurotransmitter release (Petersen et al. 1997, Frank et al.

2006). This homeostatic response requires Gbb, which is not itself the acute retrograde homeostatic signal, but rather establishes the competence of motorneurons to receive the homeostatic signal (Goold and Davis 2007). A number of genetic manipulations indicate that Gbb’s roles in regulating synaptic homeostasis, basal neurotransmission, are NMJ morphology are separable.

Perhaps surprisingly, neuronal-specific Gbb rescues both synaptic homeostasis and baseline neurotransmitter release in gbb null animals. In contrast, while muscle-derived Gbb rescues synaptic homeostasis in gbb null animals, it does not significantly rescue baseline synaptic function (Goold and Davis 2007), arguing that neuronal- and muscle-derived pools of Gbb serve distinct functions.

While our data argue strongly that Cmpy antagonizes autocrine Gbb signaling in

motorneurons to restrain morphological expansion at the NMJ, it is likely that

motorneuronal Gbb has an independent role in regulating functional development

of the NMJ. In this case, the Cmpy-Gbb complex may be active and could elicit a distinct signaling outcome than the muscle-derived pool of Gbb. Physiological analyses of cmpy mutants as well as an investigation of Gbb trafficking and secretion at the NMJ in cmpy mutants should provide critical insight into this important question.

More broadly, this study bears on the regulation of signal release in neurons. By definition, neurotransmitter is released from the presynaptic compartment and received by neurotransmitter receptors on the postsynaptic side. However, signaling pathway activity is not circumscribed in this way, and may occur at short or long range at multiple subcellular positions. Hence, neurons are likely to possess fine regulatory mechanisms controlling the release of, and response to, extracellular cues. The present work provides insight into the regulation of signaling molecules in neurons and suggests that we are only beginning to uncover mechanisms controlling signaling specificity in the developing nervous system.

Figure 2.1

Figure 2.1: CG13253 expression analysis and allele generation.

(A-D) Wild type (Oregon R) embryonic VNCs labeled with indicated markers. (A)

CG13253 mRNA is expressed in a segmentally repeating “V” shaped pattern at stage 14, similar to the expression profile of Hb9 (B). (C) CG13253 mRNA

(purple) is co-expressed with a subset of Hb9+ cells (Chesnutt et al.), the boxed area is shown at a higher magnification to the right. The double-labeled cells correspond to the cluster of RP motorneurons, including RP1, 3, 4, and 5. (D) At stage 16, CG13253 mRNA is widely expressed in the VNC. (E) Predicted domain structure of the CG13253 protein product; TM, transmembrane domain; CRR, cysteine-rich repeat; IGFBP, insulin-like growth factor binding protein-like domain;

R-K, arginine/lysine-rich domain. The predicted CRR and IGFBP domains are overlapping. (F) Genomic organization of the CG13253 locus at 77E3 on chromosome 3L. The CG13253 coding sequence is indicated in grey, and the mRNA is shown below; black and red boxes are untranslated and coding regions, respectively, and introns are depicted as thin black lines. Inverted triangles indicate the piggyBac elements utilized to generate the deletion allele, cmpy∆8.

Green arrows show the location of genomic primers used to verify deletion. (G)

PCR using the primers in (F) verifies deletion of most of the coding region of

CG13253. Amplification in a wild-type, non piggyBac-containing background yields an 8.89 kb product, while amplification across the hybrid piggyBac that remains following recombination yields a 7.86 kb product. (H) CG13253 mRNA is not expressed in the VNC of homozygous mutant embryos. Scale bars 20µm;

10µm in the right panel of (C).

Figure 2.2

Figure 2.2: cmpy functions in motorneurons to attenuate NMJ expansion.

Representative confocal images of NMJ 6/7 (A, C, E, G) and NMJ 4 (B, D, F, H)

of indicated genotypes labeled with neuronal membrane label anti-HRP. (A, B)

Wild type corresponds to heterozygous parental piggyBac elements, PBacf01736 /

PBacf02482. (C, D) cmpy∆8 homozygous NMJs display an increase in the total

number of type I synaptic boutons. (E, F) RNAi-mediated knockdown of cmpy

mRNA in motorneurons gives statistically indistinguishable overgrowth as

observed at cmpy∆8 NMJs. (G, H) Motorneuronal overexpression of cmpy in the

cmpy∆8 background restores proper growth regulation. The rescue is complete at

NMJ 4 and partial at NMJ 6/7. (I, J) Quantification of the mean number of type I boutons per indicated genotype at NMJ 6/7 and NMJ 4. Number of NMJs scored per genotype is indicated within the bars. (K) Mean number of satellite boutons at

NMJ 4. Scale bar: 20µm. Statistical comparisons are to wild type unless otherwise indicated. Error bars indicate SEM. Raw data are found in Table 2.1.

Figure 2.3

Figure 2.3: cmpy acts in the BMP signaling pathway upstream of the BMP

Type II receptor wit.

Representative confocal images of NMJ 6/7 (A, C, E, G, I) and NMJ 4 (B, D, F, H,

J) labeled with anti-HRP. (A, B) cmpy∆8 NMJs display an increase in type I

boutons. Loss of one copy of gbb (C,D), Mad (E,F), or wit (G,H) results in partial

suppression of the increase in bouton number in cmpy∆8 mutants. The reduction

in type I bouton number at wit mutant NMJs (G, H) matches the reduction

observed at wit cmpy double mutant NMJs (I, J). (K, L) Quantification of the

mean number of type I boutons per genotype at NMJ 6/7 and NMJ 4. Number of

NMJs scored per genotype is indicated within the bars. Scale bar: 20µm.

Statistical comparisons are to wild type unless otherwise indicated. Error bars

represent SEM. Raw data are found in Table 2.1.

Figure 2.4

Figure 2.4: Overexpression of cmpy suppresses NMJ expansion in larvae overexpressing neuronal gbb.

Representative confocal images of NMJ 6/7 (A, C, E) and NMJ 4 (B, D, F) labeled with anti-HRP. (A, B) Wild type corresponds to D42Gal4. (C, D)

Motorneuronal gbb overexpression promotes an increase in bouton number at

NMJ 6/7 and NMJ 4. (E, F) cmpy overexpression completely suppresses overgrowth in motorneurons overexpressing Gbb at NMJ 6/7 and NMJ 4. (G, H)

Quantification of the mean number of type I boutons per genotype at NMJ 6/7 and NMJ 4. The Gbb-induced overgrowth phenotype is not suppressed by co- overexpression of Gbb and LacZ. Scale bar: 20µm. Statistical comparisons are to wild type unless otherwise indicated. Error bars represent SEM. Raw data are found in Table 2.2.

Figure 2.5

Figure 2.5: Overexpression of cmpy suppresses gbb overexpression

phenotypes in the wing disc.

Representative micrographs of wings from female flies (A, C, E) and male flies (B,

D, F), anterior, up and distal, right. (A, D) Wild type corresponds to A9Gal4. (B, E)

Overexpression of Gbb using the wing imaginal disc driver A9Gal4 results in a blistering phenotype in the posterior compartment of wings in females, and in males results in a more severe phenotype in which wings are blistered and remain unfurled. (C, F) Co-overexpression of Cmpy with Gbb suppresses the blistering phenotype in females, and reduces the severity of the phenotype in males, such that wings are unfurled with mild blistering. Raw data are found in

Table 2.3.

Figure 2.6

Figure 2.6: RNAi-mediated knockdown of Gbb in motorneurons suppresses the cmpy LOF phenotype.

Representative confocal images of NMJ 6/7 (A, C, E, G, I) and NMJ 4 (B, D, F, H,

J) labeled with anti-HRP. (A, B) Wild type corresponds to UAS-gbbRNAi. (C, D)

Motorneuronal knockdown of Gbb by RNAi does not affect bouton number at

NMJ 6/7 or NMJ 4, whereas knockdown in muscle drives NMJ undergrowth (E,

F). Motorneuronal knockdown of Gbb suppresses NMJ overgrowth phenotypes observed at cmpy LOF NMJs (G-J). (K, L) Quantification of the mean number of type I boutons per genotype at NMJ 6/7 and NMJ 4. Scale bar: 20μm. Statistical comparisons are to wild type unless otherwise indicated. Error bars represent

SEM. Raw data are found in Table 2.4.

Figure 2.7

Figure 2.7: Cmpy physically interacts with the Gbb precursor protein.

(A) Immunoprecipitation from S2R+ cell lysates demonstrates that C-terminal tagged Cmpy-Flag and Gbb-HA fusion proteins form a complex. Top panel, both anti-Flag (Cmpy) and anti-HA (Gbb) precipitate full length, unprocessed Gbb (55 kDa). Bottom panel, anti-Flag antibody (Cmpy) precipitates both full length Cmpy

(33 kDa) and a processed, smaller Cmpy form (25 kDa). Anti-HA (Gbb) precipitates only the smaller Cmpy isoform. (B) Domains of Cmpy and Gbb used to analyze interaction by yeast two-hybrid. Gbb-Pre is the precursor form of Gbb including the prodomain and the mature ligand, Gbb-Mat. Cmpy-N is the region of Cmpy N-terminal to the transmembrane domain (black), and Cmpy-C is the region of Cmpy C-terminal to the transmembrane domain, including most of the

CRR (yellow), the IGFBP domain (magenta), and the lysine/arginine-rich region

(purple). The blue arrow indicates the approximate region of the proposed proteolytic processing of Cmpy indicated by the molecular weight of the smaller

Cmpy isoform in (A). (C) Yeast two-hybrid interactions demonstrate a physical interaction between Cmpy and Gbb. Gbb-Mat and Gbb-Pre are fused to the

GAL4 DNA binding domain (bait), while Cmpy-N and Cmpy-C are fused to the

GAL4 activation domain (prey). Cmpy-C interacts strongly with Gbb-Pre and weakly with Gbb-Mat, suggesting that Cmpy-C primarily associates with the prodomain of Gbb. Cmpy-N does not interact with either region of Gbb in this assay. (D) Yeast two-hybrid analysis demonstrates that the C-terminal portion of

Cmpy containing the lysine/arginine-rich region (Cmpy-RK), but not the CRR region (Cmpy-CR), interacts with Gbb-Pre. Scale bar in (A) is 10μm.

Table 2.1 cmpy loss-of-function and BMP genetic interaction phenotypes at the NMJ.

Table 2.2 Gbb and Cmpy gain-of-function phenotypes at the NMJ.

Table 2.3 Gbb and Cmpy gain-of-function phenotypes in the wing.

Table 2.4 gbb RNAi phenotypes at the NMJ.

CHAPTER 3:

Crimpy Sorts a BMP into the Regulated Secretory Pathway for Activity-

Dependent Release in Drosophila Motorneurons

(in preparation)

Abstract

Retrograde BMP signaling at the Drosophila neuromuscular junction (NMJ) drives morphological expansion of the presynaptic axon terminal, coupling motorneuron growth to that of the postsynaptic muscle in order to ensure balanced synaptic transmission. However the relevant BMP ligand, Glass bottom boat (Gbb), is produced by both the pre- and postsynaptic cells, raising important questions of signaling directionality specificity and the activity of motorneuronal

Gbb at the NMJ. We previously established that Crimpy (Cmpy) confers retrograde signaling directionality of pro-growth BMP signaling at the NMJ by inhibiting autocrine, pro-growth signaling in motorneurons. Yet motorneuronal

Gbb has been strongly implicated in modulating synaptic transmission at the

NMJ. The mechanisms distinguishing motorneuronal, synaptic function- promoting ligand from muscle-derived, pro-growth ligand have not been examined.

Here we demonstrate that Cmpy sorts Gbb into dense core vesicles of the regulated secretory pathway for activity-dependent secretion in Drosophila motorneurons. Gbb localizes to the NMJ when overexpressed in motorneurons, and Cmpy is necessary and sufficient for this presynaptic localization. Gbb and

Cmpy colocalize with each other, and both colocalize with a marker for dense core vesicles of the regulated secretory pathway, at presynaptic terminals and in discrete puncta within motor nerves. Gbb is secreted from motorneuron terminals in response to synaptic activity, and the activity-dependent secretion of Gbb at

the NMJ depends upon Cmpy. Furthermore, despite morphological overgrowth at

cmpy mutant NMJs, evoked synaptic transmission is impaired. We propose a model in which Cmpy sorts Gbb into dense core vesicles of the regulated secretory pathway in larval motorneurons to promote its activity-dependent secretion. We suggest that activity-dependent motorneuronal Gbb signaling modulates synaptic transmission at the NMJ, and that in the absence of this form of BMP signaling cmpy mutant NMJs, evoked synaptic transmission is impaired.

Introduction

Neural circuits integrate experience and store information by the formation and remodeling of synapses. Defining the mechanisms and molecular cues that drive such plasticity has been a major focus of contemporary neuroscience. The neurotrophin hypothesis holds that synaptic activity regulates neurotrophic growth factor expression and secretion at vertebrate central synapses, and that activity-dependent neurotrophic signaling modulates synaptic transmission and connectivity (Schinder and Poo 2000, Poo 2001). Similarly, synaptic activity regulates neuronal TGFβ growth factor secretion (Lacmann et al. 2007), and

TFGβ signaling, including BMP signaling, modulates synaptic transmission

(Krieglstein et al. 2011). Thus neurotrophins, and likely TGFβs, mediate activity- dependent changes in synaptic function and connectivity. The mechanisms that allow dual functions of growth factor signaling—such as activity-dependent synaptic modulation and classic neurotrophic activity, like neuronal survival— within a given neuronal population are not well defined.

100

The classic neurotrophins NGF, NT3, and NT4/5 largely undergo constitutive

secretion through the constitutive secretory pathway (CSP), however BDNF

primarily undergoes regulated secretion in response to activity via the regulated secretory pathway (RSP) (Haubensak et al. 1998, Thomas and Davies 2005).

The membrane-associated sorting receptors carboxypeptidase E (CPE) and

Sortilin interact with precursor BDNF in cortical and hippocampal neurons,

respectively, and deliver BDNF into the RSP for activity-dependent secretion

(Chen et al. 2005, Lou et al. 2005). Less is understood about the regulated secretion and function of activity-dependent TGFβ ligands. An appealing

hypothesis is that dedicated TGFβ sorting receptors similarly deliver precursors of activity regulated TGFβ ligand into the RSP, although no TGFβ sorting receptors have been identified to date.

Many of the key molecular players that remodel synapses and regulate synaptic transmission are evolutionarily conserved, making it likely that elucidating signaling pathways that regulate synaptic plasticity in model organisms like

Drosophila will shed light on pathway action in vertebrates. The Drosophila neuromuscular junction (NMJ) is an excellent model to investigate the molecular mechanisms underlying synaptic plasticity, since it is easily accessible, highly morphologically stereotyped, and glutamatergic, rendering it more similar to central synapses than the NMJ of vertebrates (Collins and DiAntonio 2007).

Additionally, the Drosophila NMJ is highly plastic. Many complex pathways are in place to restrict the range of synaptic activity at the NMJ so that muscles contract

101

properly (Turrigiano 1999). Of note, a BMP signaling pathway orchestrates

morphological motoraxon arbor expansion in order to accommodate a 100-fold increase in muscle surface area and preserve synaptic input during larval

development (Atwood et al. 1993). The same pathway also promotes basal

synaptic transmission (McCabe et al. 2003, Goold and Davis 2007) and mediates

homeostatic increases in neurotransmitter release in response to decreased

postsynaptic receptor activation (Goold and Davis 2007), demonstrating that

BMP signaling critically regulates synaptic plasticity at the Drosophila NMJ.

In the BMP signaling paradigm at the Drosophila NMJ, the proposed secretion of

the BMP ligand Glass bottom boat (Gbb) from the postsynaptic muscle scales

morphological expansion of the presynaptic terminal to growth of the muscle

(McCabe et al. 2003). Gbb binds the presynaptic type II receptor Wishful thinking

(Wit) and type I receptors, Saxophone (Sax) and/or Thickveins (Tkv) to initiate

signal transduction (Aberle et al. 2002, Marques et al. 2002, Rawson et al. 2003).

The intracellular signaling effectors consist of the R-Smad, Mothers against

Decapentaplegic (Mad) and the Co-Smad, Medea (Med) (Rawson et al. 2003,

McCabe et al. 2004), which translocate to the nucleus following phosphorylation by the activated type I receptors. Once in the nucleus, the Mad-Med complex associates with co-activators and co-repressors to bring about changes in gene transcription that are thought to drive a general motorneuron terminal growth promoting program. The axon terminal is expanded by the addition of synaptic boutons at the NMJ.

102

Beyond regulating morphological expansion, Gbb signaling through the aforementioned pathway components is necessary for basal synaptic transmission at the NMJ, in addition to maintaining synaptic homeostasis as previously mentioned (Goold and Davis 2007). Rescue studies demonstrated

that neuronal Gbb ligand is essential for baseline synaptic transmission at the

NMJ, while muscle-derived ligand rescues morphological growth (McCabe et al.

2003, Goold and Davis 2007). Motorneuronal Gbb also strengthens synaptic

transmission at central synapses between postsynaptic motorneurons and their

presynaptic interneuron partners (Baines 2004). Furthermore, Gbb signaling

through the type II receptor Wit in a non-canonical pathway is essential for

synapse stabilization at the NMJ (Eaton and Davis 2005), yet the source of this

ligand has not been clarified. These studies emphasize broad differences in BMP

pathway action at the NMJ, brought about by both canonical and non-canonical

BMP signal transduction cascades. Neuronal Gbb regulates synaptic function,

while muscle-derived ligand drives morphological growth. Yet the mechanisms that discriminate BMP signaling between these distinct pools of ligand to allow

such divergent signaling outcomes in motorneurons are unknown.

We previously found that the neuronal transmembrane protein Crimpy (Cmpy)

regulates BMP signaling in motorneurons to enable proper growth control at the

NMJ (James and Broihier 2011). We suggested that presynaptic expression of

Cmpy prevents autocrine pro-growth signaling in motorneurons, establishing the

103

retrograde directionality of pro-growth ligand at the NMJ. However the

mechanism by which Cmpy prevents pro-growth signaling and concurrently

allows pro-neurotransmission Gbb signaling in motorneurons was unclear. In this

study we propose that Cmpy is a sorting receptor for Gbb, delivering

motorneuronal Gbb into dense core vesicles of the RSP. We present evidence

that Cmpy is necessary and sufficient for localization of presynaptic Gbb at the

NMJ, and that Gbb secretion from motorneuron terminals is regulated by

synaptic activity. Cmpy is necessary activity-dependent release of Gbb at the

NMJ. Moreover, synaptic transmission is impaired at cmpy mutant NMJs,

suggesting that activity-dependent Gbb secretion from motorneuron terminals

regulates neurotransmission.

Material and Methods

Fly stocks

Stocks used in this work include: D42Gal4 (A. DiAntonio), UASgbb1xHA (K.

Wharton), OK6Gal4;UASanfMORG and w;;UASanfMORG (E. Levitan),

UAScmpy(7) and cmpy∆8 (James and Broihier 2011), and A9Gal4 (K. O’Conner-

Giles). UAScmpyVenus transgenic flies were generated by BestGene, Inc. All other stocks were obtained from Bloomington Stock Center.

Plasmids

EST cmpy clone (RE53920), pAWF C-terminal 3X Flag tag, and pTWV C- terminal Venus tag, Gateway System (Invitrogen) compatible vectors were

104

obtained from the Drosophila Genomics Resource Center (DGRC). Mad-Flag plasmid for the BMP signaling assay was gifted by M. Serpe (NIH). To assay

Gbb processing in S2R+ cells, either 1µg Gbb-1xHA (K. Wharton) and 1µg p8HCO control (DGRC), or 1µg Gbb-1xHA and 1µg Cmpy-Flag (James and

Broihier 2011) plasmids were transiently transfected by FuGENE HD transfection reagent (Roche). After three days, cells were processed for Western blot as previously described (Weng et al. 2011). Membranes were probed with anti-HA

Peroxidase (clone 3F10, Roche Applied Science) at 1:5000 or anti-Flag M2

(Sigma-Aldrich) at 1:2000 to detect Gbb-1xHA and Cmpy-Flag expression, respectively. Species specific secondary antibodies were used at 1:10000.

Immunoprecipitation

Larval CNS lysates for immunoprecipitation were collected as previously

described (Nechipurenko and Broihier 2012). Briefly, the CNS of 15 third instar

larvae (brain and ventral ganglion) were extracted into Triton-X lysis buffer [50

mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Triton X-100] and 1× protease inhibitor

(Roche) on ice and homogenized. Homogenized lysates were incubated on ice

for 15 minutes and cleared by centrifugation. Lysates were then incubated with

mouse anti-Gbb (Developmental Studies Hybridoma Bank) pre-bound to protein

G beads overnight at 4°C with gentle rotation. Beads were washed three times

with cold lysis buffer. Proteins were eluted in 2x sample buffer and were fractionated and subjected to Western blot as previously described. Membrane was probed with mouse anti-Gbb at 1:1000.

105

BMP Signaling Assay

To collect conditioned media for the BMP signaling assay, S2R+ cells were

transiently transfected with either 3μg Gbb-1xHA and 3μg control p8HCO

plasmids, or 3μg Gbb1xHA and 3μg Cmpy-Flag plasmids as previously

described. After two days, serum-containing media was replaced with serum-free

media. Serum-free conditioned media was collected after 12 hours and was

cleared by centrifugation. Cells were lysed and Gbb-1xHA and Cmpy-Flag fusion

protein expression in lysates was confirmed by Western blot (data not shown)

prior to using the conditioned media in the signaling assay. Membranes were

probed with anti-HA Peroxidase (clone 3F10, Roche Applied Science) at 1:5000

or anti-Flag M2 (Sigma-Aldrich) at 1:2000 to detect Gbb-1xHA and Cmpy-Flag

expression, respectively. To test for secreted BMP activity in conditioned media,

cells were transiently transfected with 2μg Mad-Flag plasmid. After three days,

the cells were collected and split evenly into three samples. Cells were treated

with (1) 100μl Gbb-1xHA conditioned media, raised to a final volume of 500μl

with serum-free media; (2) Gbb-1xHA/Cmpy-Flag conditioned media, raised to a final volume of 500μl with serum-free media, or (3) 500μl serum-free media as a negative control for 3.5 hours at room temperature with gentle rotation. The cells were cleared by centrifugation, lysed, and the supernatants were subjected to

Western blot as previously described. Membranes were probed with rabbit anti-

phospho-Smad1/5 antibody (Cell Signaling) at 1:200 to assay phosphorylation of

Mad-Flag. To control for protein loading, membranes were probed with mouse

anti-Flag M2 (Sigma-Aldrich) at 1:2000.

106

Immunohistochemistry

For all larval experiments, ten virgin females were crossed to five males, and

bottles were maintained at 25°C for 4 days before removing adults. Dissection of

wandering third instar female larvae was carried out in ice-cold PBS, and body

walls were fixed in Bouin’s Fixative (Polysciences). The following primary

antibodies were used: rat anti-HA High Affinity (clone 3F10, Roche Applied

Sciences) at 1:100, chicken anti-GFP (Abcam) at 1:400, rabbit anti-RFP (Abcam) at 1:400, DyLight-594 anti-HRP (Jackson ImmunoResearch) at 1:500, and rabbit anti-pMad (kindly provided by Carl-Henrik Heldin). The following species-specific

secondary antibodies were used: Alexa Fluor 488, Alexa Fluor 568, and Alexa

Fluor 647 (Invitrogen) at 1:300. Body walls were incubated in primary antibody

overnight at 4°C or for 2-3hrs at room temperature, and in secondary antibody for

2 hours at room temperature.

FM1-43 and FM4-64 dye uptake assays and live imaging

FM dye loading was carried out as previously described (Nechipurenko and

Broihier 2012). Briefly, wandering L3 larvae were dissected in HL3 without Ca2+

(110mM NaCl, 5mM KCl, 10mM NaHCO3, 5mM HEPEs, 30mM sucrose, 5mM

trehalose, and 10mM MgCl2). Recycling synaptic vesicles were labeled by

incubating dissected body walls in 90mM K+ Jan’s saline (45mM NaCl, 90mM

KCl, 2mM MgCl2, 36mM sucrose, 5mM HEPES, and 2mM CaCl2, pH 7.3)

containing 4μM FM1-43 for 5 minutes or 4μM FM4-64 dye for 2 minutes

(Invitrogen). Body walls were washed with Ca2+ free HL3 prior to imaging. For

107

FM4-64 colocalization with Cmpy-Venus in live tissue, body walls were spread in

Vectashield mounting medium (Vector Laboratories) and immediately imaged on

an LSM 510 Meta laser-scanning system (Carl Zeiss) confocal microscope at

100X. For FM1-43 experiments, labeled vesicles were imaged on a Zeiss

Axioplan 2 (Carl Zeiss) at 40X.

High K+ depolarization and channelrhodopsin-2 light activation paradigms

To assay changes in intracellular Gbb-HA levels following high K+ depolarization,

UASgbb1xHA was expressed in larval motorneurons using D42Gal4. Larvae were dissected in ice cold PBS and incubated in 90mM K+ Jan’s saline (see

above) for five minutes at room temperature. Controls were incubated in normal

Jan’s saline, and to probe Ca2+ dependence of changes in intracellular Gbb-

1xHA, body walls were incubated in modified 90mM K+ Jan’s saline without Ca2+.

Body walls were washed in PBS and fixed in Bouin’s fixative. Equal numbers of

body walls from each condition were dissected and processed for

immunohistochemistry within the same tube, to control for staining variability.

As an independent measure of activity-dependent changes in intracellular Gbb-

1xHA levels, the blue light-activated cation channel, channelrhodopsin-2 was

used to drive synaptic activity. UASChR2 and UASgbb1xHA transgenes were expressed in larval motorneurons using D42Gal4 and levels of Gbb-HA were

assayed following exposure of body walls to blue light through a Zeiss Axioplan 2

band pass GFP filter. Larvae were reared on 100μM all-trans retinal (Sigma-

108

Aldrich) containing molasses caps covered in 100μM all-trans retinal yeast paste

at 25°C in the dark. As proof of principle that fluorescence through the GFP filter

activates ChR2, ElavGal4 was used to drive ChR2 in all post-mitotic neurons and contractions of third instar larvae reared on all-trans retinal food in response to blue light were assayed as previously described (Schroll et al. 2006). To

measure changes in intracellular Gbb-HA following blue light stimulation, third

larvae were dissected in ice cold PBS. PBS was replaced with room temperature

Jan’s saline and body walls were exposed to blue light for 5 minutes and then

fixed in Bouin’s fixative. Control and test body walls were processed for

immunohistochemistry in the same tube to control for staining variability.

Electrophysiology

Intracellular recordings were obtained from muscle 6, segment A2–A4 in third

instar larvae. Larvae were initially dissected in HL3 without Ca2+, and prior to

recording the saline was replaced with1.0mM Ca2+ HL3. All recordings were

obtained from female larvae. Data was collected from cells with an input

resistance greater than 4 MΩ and resting membrane potential less than -55mV.

Sharp electrodes of a resistance between 8–20 MΩ, filled with 3M KCl, were used. Recordings were performed using a Multiclamp 700A (Axon Instruments), and data was filtered at 2 kHz and digitized using a Digidata 1322A (Axon

Instruments). Stimulation was carried out with a Master-8 Stimulator (A.M.P.I.) at

0.2 Hz. Care was taken to ensure that both motorneurons innervating muscle 6 were recruited. MiniAnalysis software (Synaptosoft, Inc.) was used to calculate

109

mEJP and EJP amplitudes. mEJPs were analyzed over a one minute interval, and 20 consecutive EJPs were averaged per cell. The mean EJP amplitude was divided by the mean mEJP amplitude to calculate quantal content.

Imaging and data analysis

Larval NMJs were imaged on a Zeiss Axio Imager.ZI confocal microscope at 63X and 100X, and larval motor nerves were imaged at 100X with 7X zoom.

Brightness and contrast were adjusted in Adobe Photoshop CS5. For all experiments comparing levels of fluorescence, larval body walls were stained in the same tube, and control and test images were processed similarly, if at all. For transgene control experiments (Table 3.1), quantification of type I glutamatergic boutons was carried out as previously described at NMJ 6/7 and NMJ 4 (James and Broihier 2011). For data analysis, groups of means were compared by one- way ANOVA, and pairs of means by the unpaired Student’s t-test.

110

Results & Discussion

Crimpy is dispensable for Gbb processing and secretion

We previously found that Cmpy restrains autocrine Gbb signaling in Drosophila motorneurons to enable proper motoraxon elaboration at the NMJ during larval development (James and Broihier 2011). Although Cmpy restrains motorneuronal pro-growth Gbb signaling activity, motorneuronal Gbb regulates synaptic transmission at central synapses (Baines 2004) and neuronal Gbb is necessary for synaptic function at the NMJ (McCabe et al. 2003, Goold and

Davis 2007). Thus motorneurons are poised to respond to two pools of Gbb: the retrograde pool derived from the muscle that scales presynaptic growth to postsynaptic growth; and the neuronal pool that modulates synaptic transmission. The mechanism by which Cmpy restrains autocrine, pro-growth

Gbb signaling activity yet allows Gbb signaling from motorneurons to regulate synaptic transmission is unclear. We imagined that Cmpy could distinguish the neuronal pool of Gbb that is necessary for basal neurotransmission from the pro- growth, retrograde pool of ligand at the NMJ by altering processing of motorneuronal Gbb, effectively changing its appearance at the NMJ. Consistent with this hypothesis, a novel furin convertase processing site in precursor Gbb generates a biologically active, large Gbb ligand with different signaling capacity than the canonical mature ligand (Akiyama et al. 2012).

To investigate if Cmpy alters Gbb processing, we initially assayed ligand processing in S2 cells in the presence and absence of Cmpy. We and others

111

have found that large N- and C-terminal epitope tags like GFP impair processing

of Gbb precursor protein, likely by altering ligand folding and occluding

convertase cleavage sites, thereby diminishing signaling capacity of tagged Gbb

fusion proteins (data not shown) (Akiyama et al. 2012). Using a Gbb-1xHA plasmid with a single internal hemagglutinin (HA) epitope that is processed and retains signaling capability comparable to untagged Gbb (Akiyama et al. 2012), we detected precursor Gbb (Gbb-Pre, 55 kDa) and processed, mature peptide

(Gbb-Mat, 17 kDa) in S2 cell lysates using an anti-HA antibody (Fig. 3.1A).

Ligand processing was unchanged by co-expression of Cmpy in S2 cells (Fig.

3.1A). To test if Cmpy alters Gbb processing in vivo, we took advantage of a newly generated antibody that biochemically detects endogenous Gbb (Akiyama et al. 2012). Using this antibody we immunoprecipitated endogenous Gbb from ventral ganglia of wild type and cmpyΔ8 third instar larvae. Both Gbb-Pre and

Gbb-Mat were detected in wild-type ventral ganglia, and processing was unchanged in cmpyΔ8 mutants (Fig. 3.1B). Gbb immunoprecipitated from the body wall was comparably processed (data not shown). These data indicate that

Cmpy does not alter Gbb processing.

We previously observed that a C-terminal Arg-Lys (R-K) rich region of Cmpy

binds to precursor Gbb in vitro (James and Broihier 2011), reminiscent of

interactions between the intracellular BMP antagonist Gremlin and BMP4.

Gremlin binds to precursor BMP4 through an R-K rich region, and this interaction

prevents BMP4 secretion (Sun et al. 2006). We wondered if Cmpy allows sole

112

reception of pro-growth Gbb at the NMJ by preventing the secretion of

motorneuronal Gbb in the periphery. To test this hypothesis, we turned to a well-

established in vitro BMP signaling assay (Ross et al. 2001) to ask if Cmpy

prevents secretion of functional Gbb ligand from S2 cells. S2 cells expressing a

Mad-Flag fusion protein will phosphorylate Mad-Flag in response to application of

exogenous BMP ligand (Ross et al. 2001, Serpe et al. 2008). To ask if Cmpy

prevents Gbb secretion in S2 cells, we assessed signaling activity present in conditioned media (CM) isolated from cells that expressed either Gbb alone or

Gbb in conjunction with Cmpy. We treated Mad-Flag expressing cells with Gbb-

CM or Gbb/Cmpy-CM and assayed levels of Mad-Flag phosphorylation in response to CM treatment. We found that Mad-Flag was similarly phosphorylated by both treatment paradigms (Fig. 3.1C), implying that Gbb signaling activity

secreted from S2 cells is not decreased by co-expression of Cmpy. Together

these data suggest that Cmpy does not regulate Gbb processing or secretion;

however Cmpy could still function to preclude motorneuronal Gbb secretion in

the periphery, thus permitting reception of pro-growth ligand exclusively, by

preventing Gbb trafficking to the NMJ.

Crimpy promotes localization of Gbb at presynaptic terminals

Studies examining localization of Gbb in vivo have been limited by inability to

detect endogenous protein. As mentioned previously, tagged Gbb variants are

not processed properly, which confounds interpretation of localization studies

using tagged transgenes. To overcome this limitation, we evaluated localization

113

of Gbb-HA fusion protein expressed from a UASgbb1xHA transgene that was derived from the aforementioned Gbb-1xHA plasmid that is properly processed and retains signaling activity in vitro comparable to wild-type Gbb (Akiyama et al.

2012). Importantly, UASgbb1xHA drives NMJ overgrowth and wing blistering phenotypes comparable to wild-type Gbb transgenes when expressed in developing larval motorneurons using D42Gal4 and in the wing imaginal disc using A9Gal4, respectively (Table 3.1, data not shown).

To probe if Cmpy prevents peripheral Gbb secretion by excluding motorneuronal

Gbb from the NMJ, we first expressed UASgbb1xHA in larval motorneurons with

D42Gal4 and evaluated Gbb-HA localization using antibodies against the HA epitope. Gbb-HA protein is expressed in motorneuron cell bodies in the ventral ganglion (data not shown) and is present in presynaptic terminals, displaying a punctate localization pattern (Fig. 3.2A). We reasoned that if Cmpy normally prevents Gbb trafficking to the NMJ, Gbb-HA localization at the NMJ would be increased in cmpyΔ8 mutants. Surprisingly, Gbb-HA levels at the NMJ are decreased 2-fold in cmpyΔ8 mutants (Fig. 3.2B, C), demonstrating that Cmpy is necessary for Gbb trafficking to the NMJ.

To ask if Cmpy is also sufficient for presynaptic Gbb-HA localization we co- overexpressed UASgbb1xHA and UAScmpyVenus transgenes in larval motorneurons using D42Gal4 and evaluated presynaptic levels of Gbb-HA. Of note, UAScmpyVenus, like wild-type Cmpy transgenes, does not affect NMJ

114

growth when overexpressed in larval motorneurons in an otherwise wild-type

background, and completely rescues overgrowth at cmpyΔ8 NMJs (Table 3.1).

Remarkably, presynaptic Gbb-HA levels are increased 2.5-fold by co- overexpression of Cmpy-Venus with Gbb-HA (Fig. 3.2 D-F). These data argue against a role for Cmpy in preventing Gbb trafficking to the NMJ, and imply that

Cmpy actually promotes Gbb delivery to the NMJ, although these data do not rule out the possibility that Cmpy retains Gbb at the NMJ following its delivery.

Gbb and Crimpy colocalize in vivo and are found in dense core vesicles

Since Cmpy and Gbb associate in vitro (James and Broihier 2011) and Cmpy-

Venus enhances presynaptic Gbb-HA localization, we imagined that Cmpy could function as a neuronal adaptor to deliver Gbb to axon terminals. Or as

mentioned, Cmpy could retain Gbb-HA at the NMJ post-delivery. In either course,

we predicted that Gbb-HA and Cmpy-Venus would colocalize within synaptic boutons. To conduct colocalization experiments, we first characterized Cmpy-

Venus expression in larval motorneurons using D42Gal4. Cmpy-Venus is

expressed in motorneuron cell bodies of the ventral ganglion (data not shown)

and at the NMJ, where it is found exclusively in the presynaptic compartment

(Fig. S3.1). Cmpy-Venus is detected using an anti-GFP antibody that is specific

to expression of the tagged transgene, as revealed by the lack of anti-GFP labeling the NMJ in the UAScmpyVenus background (Fig. S3.1A, top). Cmpy-

Venus partially colocalizes with the neuronal membrane marker HRP at the cell

115

surface, but does not overlap with the postsynaptic subsynaptic reticulum marker, Dlg (Fig. S3.1A, B).

When we expressed both transgenes in larval motorneurons using D42Gal4 and evaluated their localization at the NMJ, we observed marked overlap of Gbb-HA with Cmpy-Venus within synaptic boutons (Fig.3.3A). If Cmpy delivers Gbb to

NMJ, we reasoned that Cmpy-Venus and Gbb-HA expression would also overlap within larval motor nerves. Cmpy-Venus in motoraxons of larval nerves is observed in discrete puncta reminiscent of trafficking vesicles (Fig. 3.3B, top).

Gbb-HA was also found in puncta within motoraxons (Fig. 3.3B, middle), where it colocalizes with Cmpy-Venus (Fig. 3.3B, bottom). This is consistent with the idea that Cmpy delivers Gbb to the presynaptic terminal, as opposed to simply retaining Gbb post-delivery.

Two types of vesicles are present within synaptic terminals at the NMJ: small, clear neurotransmitter-containing vesicles and large dense core vesicles (DCVs).

In our initial characterization of Cmpy-Venus expression at the NMJ, we did not observe considerable overlap between Cmpy and the glutamate transporter

DVGLUT (Fig. S3.1C), a marker for glutamate-containing synaptic vesicles

(Daniels et al. 2004), suggesting that Cmpy is not present in recycling synaptic vesicles at the NMJ. Neuropeptides and growth factors are often sorted into large

DCVs of the regulated secretory pathway (RSP) and undergo Ca2+ dependent,

regulated exocytosis in response to neuronal activity (Zhang et al. 2010,

116

Vazquez-Martinez et al. 2012). To ask if Cmpy is sorted into DCVs in larval

motorneurons, we examined colocalization between Cmpy-Venus and atrial natriuretic factor (Anf), a dense core vesicle marker (Rao et al. 2001), tagged with mOrange. Cmpy-Venus expression overlaps with Anf-mOrange at the NMJ

(Fig. 3.3C) and in discrete puncta within motor nerves (Fig. 3.3D). Additionally,

Gbb-HA expression also overlaps with Anf-mOrange in the presynaptic terminal

(Fig. 3.3E) and in motoraxons (Fig. 3.3F).

To ask if Cmpy-Venus is sorted into DCVs specifically as opposed to small, clear synaptic vesicles, we examined colocalization between Cmpy-Venus and FM4-

64 in live tissue. FM4-64 is a red fluorescent lipophilic dye that labels neurotransmitter-containing recycling synaptic vesicles, but not DCVs (Levitan

2012), in presynaptic terminals following high K+ depolarization. Individual Cmpy-

Venus puncta are readily observed in live tissue at the NMJ, and following 5

minutes of 90mM K+ stimulation, there is very little overlap of Cmpy-Venus with

FM4-64 labeled recycling synaptic vesicles (Fig. 3.3G), further supporting the

hypothesis that Cmpy-Venus is sorted into DCVs of the RSP. However a detailed

ultrastructural analysis of Cmpy-Venus and Gbb-HA localization at the NMJ is

necessary to definitively place Cmpy and Gbb in DCVs exclusively.

Devoted molecular chaperones in the endoplasmic reticulum recognize specific

sorting signals within nascent proteins to actively sort and concentrate secretory

cargo destined for further processing in the Golgi complex (Vazquez-Martinez et

117

al. 2012). Since this is a tightly regulated, active process, it is unlikely that the

proposed sorting of Cmpy-Venus and Gbb-HA into DCVs of the regulated

secretory pathway is an artifact of the overexpression system. Together these

data indicate that Cmpy-Venus and Gbb-HA are sorted into DCVs, but not small

clear core vesicles, for regulated secretion from larval motorneurons.

Activity regulates levels of Gbb-HA at the larval NMJ

Since Cmpy-Venus and Gbb-HA colocalize with a marker for DCVs at the NMJ,

and neuropeptides that are sorted into DCVs undergo Ca2+ dependent, regulated

exocytosis in response to synaptic activity (Zhang et al. 2010), we predicted that activity would regulate presynaptic Gbb-HA levels. To test this hypothesis, we stimulated motorneurons by incubating larval body wall preparations with 90mM

K+ saline for 5 minutes and then assessed the amount of Gbb-HA present in

presynaptic terminals following stimulation compared to unstimulated controls.

After stimulation, Gbb-HA levels at the NMJ are decreased by 60% compared to unstimulated body walls (Fig. 3.4A, B), implying that Gbb-HA is secreted in response to electrical activity. Gbb-HA levels were normalized to levels of HRP as an internal control for staining variability.

Secretion of DCVs, like small, clear neurotransmitter-containing synaptic vesicles, is dependent upon Ca2+ influx following an action potential. To support

that Gbb-HA is secreted in a Ca2+ dependent manner in response to synaptic activity, we incubated body walls in high K+ saline without Ca2+ and evaluated

118

Gbb-HA levels at the NMJ. Without Ca2+ in the external solution, Gbb-HA levels

were unchanged at the NMJ (Fig. 3.4B). Interestingly, eliminating extracellular

Ca2+ significantly increased Gbb-HA levels beyond those observed in

unstimulated controls (Fig. 3.4B). This likely reflects the elimination of

spontaneous activity in the absence of Ca2+, signifying that spontaneous activity may regulate growth factor secretion at the NMJ.

As an independent measure of the activity-dependence of Gbb-HA localization in presynaptic terminals, we induced synaptic activity in larval motorneurons using the blue light-activated cation channel, Channelrhodopsin-2 (ChR2). We exposed body wall preparations expressing UASChR2 and UASgbb1xHA transgenes expressed in larval motorneurons with D42Gal4 to blue light for 5 minutes. Following blue light-activated depolarization, Gbb-HA levels were decreased in presynaptic terminals by 43% compared to controls (Fig. 3.4C).

This is consistent with decreases in Gbb-HA following high K+ stimulation.

Following stimulation by either method, we do not detect a shift of Gbb-HA protein into motor nerves, suggesting that Gbb-HA is not trafficked away from the

NMJ in response to synaptic activity (data not shown). Together these experiments strongly indicate that Gbb-HA is secreted from motorneuron terminals in response to synaptic activity, although we cannot rule out the possibility that Gbb-HA is degraded in response to activity.

119

In addition to neuropeptides, neurotrophic growth factors like BDNF may also be

secreted from DCVs in response to electrical activity (Haubensak et al. 1998, Lu

2003, Dieni et al. 2012). Indeed, BDNF preferentially undergoes activity-

dependent secretion through the RSP (Lu 2003, Thomas and Davies 2005).

BDNF delivery into the RSP depends upon interactions with the sorting receptors

Sortilin and Carboxypeptidase E (Chen et al. 2005, Lou et al. 2005). Since

growth factors signal widely outside of the nervous system and through the

constitutive secretory pathway (CSP) in addition to the RSP, it is reasonable to

predict that like BDNF, other growth factors that undergo regulated secretion rely

on cell type specific sorting receptors to direct them into the RSP. We therefore

reasoned that Cmpy is a sorting receptor that delivers Gbb into the RSP within

larval motorneurons. To test this, we examined activity-dependent regulation of

Gbb-HA levels within presynaptic terminals following high K+ depolarization in a

cmpyΔ8 homozygous mutant background. Gbb-HA levels at the NMJ are

unchanged following stimulation in a cmpyΔ8 mutant background (Fig. 3.4D).

From this we conclude that cmpy is necessary for activity-dependent secretion of

Gbb-HA at the NMJ. Since the default method of secretion for regulated

secretory cargo that is not properly targeted into the RSP is the CSP (Burgess and Kelly 1987), we propose that overgrowth of cmpyΔ8 terminals is due to

ectopic constitutive secretion of motorneuronal Gbb at the NMJ.

Cmpy is necessary for proper synaptic transmission at the NMJ

Activity-dependent secretion of growth factors like BDNF is critical for regulation

120

of synaptic plasticity (Schinder and Poo 2000, Lu 2003). In particular, acute

BDNF application facilitates long-term potentiation (LTP) at hippocampal synapses and both BDNF and NT3 rapidly augment neurotransmitter release at the vertebrate NMJ in vitro (Lu 2003). Further, BMP signaling has been implicated in enhancing presynaptic neurotransmitter release and LTP at central synapses in vertebrates (Krieglstein et al. 2011). In light of these studies, the presynaptic requirement of Gbb for basal neurotransmission at the NMJ

(McCabe et al. 2003, Goold and Davis 2007), and sorting of Gbb-HA into the

RSP for activity-dependent secretion, we hypothesized that motorneuronal Gbb

signaling is distinguished from pro-growth Gbb signaling at the NMJ by its link to

synaptic activity. If Cmpy sorts motorneuronal Gbb into the RSP for activity-

dependent secretion, and this determines a role for motorneuronal ligand in

regulating synaptic function as opposed to growth at the NMJ, we predicted that

neurotransmission at cmpyΔ8 NMJs would be impaired, despite a 50% increase

in the number of synaptic boutons (James and Broihier 2011).

To investigate the relationship between Cmpy and synaptic function, we

examined evoked and spontaneous neurotransmitter release at muscle 6 of

cmpyΔ8 and control body walls by intracellular voltage recording. We observed a

28% decrease in the amplitude of evoked excitatory junction potentials (EJPs) at

cmpyΔ8 NMJs compared to controls (42.4 ± 0.3 mV in wild type versus 30.7 ± 0.4

mV in cmpyΔ8; n=18) (Fig. 3.5A, B). The amplitude of spontaneous miniature

excitatory junction potentials (mEJPs) was unchanged in cmpy mutants (1.12 ±

121

0.01 mV in wild type versus 1.16 ± 0.01 mV in cmpyΔ8; n=16), as was the

frequency of mEJPs (1.52 ± 0.16 Hz in wild type versus 1.93 ± 0.23 Hz in

cmpyΔ8; n=16) in the absence of stimulation (Fig. 3.5A, C; data not shown).

Quantal content was decreased by 30% at cmpyΔ8 NMJs (Fig. 3.5D). Together these data point to a presynaptic defect in neurotransmitter release. Impaired

synaptic function at cmpy mutant NMJs is due to loss of cmpy in motorneurons specifically as presynaptic expression of a wild-type Cmpy transgene in larval motorneurons increases EJP amplitude and quantal content by 30% over cmpy mutant background controls (Fig. 3.5A, B, D). Again, mEJP amplitude and frequency was unaffected in these backgrounds (Fig. 3.5 A, C; data not shown).

To gain insight into the mechanism of presynaptic dysfunction at cmpy mutant terminals, we evaluated synaptic vesicle (SV) cycling at the NMJ using the lipophilic dye FM1-43 following 90mM K+ depolarization. FM1-43 labels recycling

SVs within the nerve terminal, therefore defects in FM1-43 labeling at the NMJ indicate a faulty SV cycle (Kuromi and Kidokoro 2005). High K+ stimulation

strongly labeled synaptic boutons at NMJ4 in controls; however, SVs are labeled

by FM1-43 30% less efficiently at cmpyΔ8 terminals (Fig. 3.5E, F).

Growth factor secretion is thought to be a driving factor behind neurotransmitter

release (Lacmann et al. 2007), at least in the context of synaptic plasticity

(Schinder and Poo 2000). We wondered if activity-dependent secretion of Gbb at the NMJ similarly enhances presynaptic release of neurotransmitter-containing

122

SVs. Indeed, impaired evoked synaptic transmission in BMP pathway and

cmpyΔ8 mutants and aberrant SV cycling at cmpyΔ8 NMJs is consistent with the idea that activity-dependent Gbb secretion from motorneurons enhances SV release. We imagined that in the absence activity-dependent Gbb secretion there might be a build-up of SVs at cmpyΔ8 terminals. We examined expression of the synaptic vesicle markers DVGLUT and synaptotagmin at cmpyΔ8 NMJs and

found nearly a 2-fold increase in levels of both (Fig. S3.2, data not shown),

consistent with impaired SV release. However, ultrastructural analysis and

quantification of small, clear cored SVs is necessary to validate an actual

increase in vesicle number at cmpy mutant terminals. Intriguingly, ultrastructural

analyses of wit and gbb mutant NMJs revealed impaired synapse stability, evidenced by presynaptic membrane detachment near active zones and floating electron dense T-bodies within the cytoplasm of synaptic boutons that cluster

SVs away from the cell membrane (Aberle et al. 2002, McCabe et al. 2003). SVs clustered at T-bodies within the cytoplasm could account for increased expression of DVGLUT and synaptotagmin at cmpyΔ8 NMJs. It will be interesting

to learn if SV-clustering T-bodies are present at cmpy mutant terminals.

We found the decrease in synaptic transmission at cmpy mutant NMJs

perplexing, since Gbb promotes synaptic transmission at the NMJ and

overgrowth at cmpy Δ8 NMJs is driven by excessive Gbb signaling (James and

Broihier 2011). To further investigate Gbb signaling in cmpy mutants, we probed expression of phospho-Mad (pMad), the downstream BMP signal effector, in

123

cmpyΔ8 motorneurons. We found that pMad expression was increased 2.7-fold at

cmpy mutant NMJs compared to wild-type controls (Fig. S3.3A, B). However

pMad accumulation in motorneuron nuclei is unaltered in cmpy mutants (Fig.

S3.3C). Accumulation of pMad at cmpyΔ8 NMJs suggests that local pMad

signaling is capable of driving overgrowth at the NMJ, and since pMad is

unaffected in cmpyΔ8 motorneuron nuclei, it implies that cmpy-dependent

modulation of synaptic transmission occurs through a non-canonical pathway.

Non-canonical Gbb signaling downstream of Wit regulates synapse stability

through LIM kinase at the NMJ, and interestingly overexpression of wild-type LIM kinase rescues defects in synaptic transmission at wit mutant NMJs (Eaton and

Davis 2005). However the source of Gbb ligand that regulates synaptic stability through LIM kinase at the NMJ remains obscure. It is tempting to speculate that cmpy-dependent delivery of motorneuronal Gbb into the RSP, and subsequent activity-dependent Gbb release at the NMJ, modulates synaptic transmission through a non-canonical pathway involving LIM kinase.

Together, the experiments presented here indicate that Cmpy delivers Gbb into dense core vesicles of the RSP for activity-dependent secretion from Drosophila motorneurons (Fig. 3.6A). We propose that, in the absence of Cmpy, Gbb protein is mis-sorted into the CSP and that constitutive release of Gbb from motorneurons drives overgrowth at cmpy∆8 NMJs (Fig. 3.6B). Furthermore, Cmpy

is necessary for proper synaptic transmission, pointing towards a role for activity-

124

dependent, motorneuronal Gbb signaling in regulating synaptic transmission at the NMJ.

Conclusions

Our previous work demonstrated that Cmpy prevents autocrine, pro-growth Gbb signaling from motorneurons at the Drosophila NMJ. Yet expression of Gbb in motorneurons and the function of neuronal Gbb activity in modulating synaptic transmission at central synapses and at the NMJ indicated the presence of a

BMP pathway distinct from the pro-growth pathway that regulates synaptic function in motorneurons. Here we investigated mechanisms to distinguish between pro-neurotransmission and pro-growth BMP signaling at the NMJ. We found that Gbb is sorted into dense core vesicles in motorneurons for regulated secretion in response to synaptic activity.

We envision a model in which activity-dependent secretion of motorneuronal Gbb at the NMJ regulates neurotransmission as opposed to morphological growth.

Since Gbb-HA is delivered less efficiently to the NMJ in cmpy mutants and activity-dependent secretion of Gbb-HA is abolished at cmpyΔ8 NMJs, an appealing hypothesis is that Cmpy-dependent sorting of Gbb into the RSP and ensuing activity-dependent Gbb secretion at the NMJ strengthens synaptic transmission. In line with this, despite a 50% increase in the number of synaptic boutons, which have an appropriate density of active zones and appear morphologically normal at the level of light microscopy (James and Broihier

125

2011), evoked synaptic transmission at cmpyΔ8 NMJs is diminished. This is consistent with the hypothesis that activity-dependent Gbb secretion strengthens synaptic transmission. However further validation of our model awaits functional rescue experiments in a gbb cmpy double mutant background. We anticipate that if Cmpy-dependent, regulated secretion of Gbb strengthens synaptic transmission at the NMJ while constitutive release through the CSP drives the addition of synaptic boutons, motorneuronal expression of Gbb in a gbb cmpy background will result in impaired rescue of synaptic function compared to motorneuronal rescue in gbb single mutants.

Our identification of Cmpy as a sorting receptor that delivers a BMP into the RSP in Drosophila motorneurons provides an important step in discerning the mechanism of activity-dependent TGFβ secretion from vertebrate neurons. Since

TGFβs are released from hippocampal neurons in response to activity and

TGFβ/BMP signaling modulates synaptic activity in the vertebrate hippocampus

(Lacmann et al. 2007, Sun et al. 2007, Krieglstein et al. 2011), it is likely that cell- type specific TGFβ/BMP sorting receptors similarly regulate pro- neurotransmission growth factor signaling in vertebrates.

126

Figure 3.1

127

Figure 3.1: Crimpy does not alter Gbb processing or secretion.

(A) Western blot from Drosophila S2R+ cell lysates demonstrates that a Gbb-

1xHA fusion protein is processed comparably in the presence or absence of

Cmpy-Flag. Both full-length, unprocessed Gbb-1xHA (Gbb-Pre, 55 kDa) and processed, C-terminal peptide (Gbb-Mat, 17 kDa) forms of ligand are present when Cmpy-Flag is co-transfected with Gbb-1xHA. (B) Immunoprecipitation of

endogenous Gbb in a cmpyΔ8 homozygous background. Anti-Gbb antibody

precipitates both Gbb-Pre (55 kDa) and Gbb-Mat (17 kDa) from 3rd instar ventral

ganglia lysates of both wild type and cmpyΔ8 larvae. Wild type corresponds to

heterozygous parental PBac elements, PBacf01736 / PBacf02482. (C) Induction of

phosphorylated-Mad (pMad) in S2 cells transfected with Mad-Flag fusion protein following treatment with conditioned media from S2 cells expressing Gbb-1xHA, or Gbb-1xHA and Cmpy-Flag. Conditioned media was collected from S2 cells transfected with equivalent amounts of Gbb-1xHA or Gbb-1xHA/Cmpy-Flag

plasmids; (-) control refers to treatment with naïve media. Mad-Flag is

comparably phosphorylated after S2 cells were treated with equivalent volumes

of Gbb-1xHA or Gbb-1xHA/Cmpy-Flag conditioned media. Bottom panel is a

loading control demonstrating equivalent levels of Mad-Flag in each condition.

128

Figure 3.2

129

Figure 3.2: Crimpy is necessary and sufficient for Gbb localization at the

NMJ.

Representative confocal images of Drosophila NMJ4 (A, B, D, E) labeled with anti-HA to mark Gbb-HA transgene expression and anti-HRP to mark the neuronal membrane. (A-C) Crimpy is necessary for Gbb-HA localization at the

NMJ. Gbb-HA expressed in motorneurons localizes to the NMJ at NMJ4, and this localization is decreased in a cmpy∆8 homozygous mutant background (B). (C)

Quantification of the ratio of Gbb-HA levels to HRP levels for the indicated

genotype. **, P<0.01. n, number of NMJs scored from 7 animals. (D-F) Crimpy is sufficient for Gbb-HA localization at the NMJ. (E) Gbb-HA levels are increased at

NMJ4 by co-overexpression of Cmpy-Venus with Gbb-HA in larval motorneurons

compared to overexpression of Gbb-HA alone (D). (F) Quantification of the ratio

of Gbb-HA levels to HRP levels for the indicated genotype. ***, P<0.001. n,

number of NMJs scored from 10 animals. Scale bars: 10μm. Data represent the

mean ± SEM.

130

Figure 3.3

131

Figure 3.3 (Continued)

132

Figure 3.3: Crimpy and Gbb colocalize within dense core vesicles of

Drosophila motorneurons.

Representative confocal images of Drosophila NMJ4 (A, C, E, G) and larval

motor nerves (B, D, F). (A, B) Cmpy-Venus (anti-GFP, green) colocalizes with

Gbb-HA (anti-HA, purple) within synaptic boutons at NMJ4 (A) and with discrete

Gbb-HA puncta in motor nerves (B). (C, D) Cmpy-Venus (green) colocalizes with the dense core vesicle marker, Anf-mORG (anti-RFP, purple) in synaptic boutons at NMJ4 (C) and within motor nerves (D). (E, F) Gbb-HA (anti-HA, green) also colocalizes with Anf-mORG at NMJ4 (E) and within larval motor nerves (F).

Together these colocalization studies indicate that Cmpy-Venus and Gbb-HA are sorted into dense core vesicles of the regulated secretory pathway and traffic within dense core vesicles to the NMJ (D, F). (G) Cmpy-Venus (green) does not strongly colocalize with FM4-64 dye (purple) labeled recycling synaptic vesicle pool in live tissue after 5 minutes high K+ depolarization. Scale bars: 5μm for (A,

C, E), 10μm for (B, D, F) and (G).

133

Figure 3.4

134

Figure 3.4 (Continued)

135

Figure 3.4: Activity regulates Gbb localization at the NMJ.

(A) Representative confocal images of Drosophila NMJ4 labeled with anti-HA to

mark Gbb-HA transgene expression, and anti-HRP to mark the neuronal

membrane. (Top) Gbb-HA localization at NMJ4 following 5 minutes incubation

with K+ free saline. (Bottom) Gbb-HA localization at NMJ4 is decreased following

5 minutes depolarization with 90mM K+ saline. Scale bar: 10μm. (B)

Quantification of the ratio of Gbb-HA levels to HRP levels for the given condition.

Decreased levels of Gbb-HA following high K+ depolarization depend on extracellular Ca2+. * P<0.05, *** P<0.001. n, number of NMJs scored from at least

8 animals from two independent experiments. (C) Quantification of the ratio of

Gbb-HA levels to HRP levels in larvae expressing Gbb-HA and the blue light-

activated channel, Channelrhodopsin-2 in motorneurons. Depolarization induced

by 5 minutes exposure to blue light significantly reduces the levels of Gbb-HA at

NMJ4. ***, P<0.001. n, number of NMJs scored from 9 animals from two

independent experiments. (D) Quantification of the ratio of Gbb-HA levels to HRP

levels in larvae expressing Gbb-HA motorneurons in a cmpy∆8 homozygous mutant background. Gbb-HA levels remain unchanged following 5 minutes of high K+ depolarization at NMJ4 in the absence of cmpy. n.s., not significant. n,

number of NMJs scored from at least 10 animals from two independent experiments. Data represent the mean ± SEM. Scale bar: 10µm.

136

Figure 3.5

137

Figure 3.5: Synaptic transmission is impaired at cmpy mutant NMJs.

(A-D) Electrophysiological analysis of cmpy mutant NMJs. (A) Representative

traces of evoked and spontaneous potentials at control, cmpy∆8, and motorneuron-rescued NMJs. Control corresponds to heterozygous parental PBac

elements, PBacf01736 / PBacf02482. (B) EJP amplitude is decreased by 28% at

muscle 6 of cmpy∆8 mutants compared to control. EJP amplitude is restored by

expression of a wild-type Cmpy transgene within motorneurons in a cmpy∆8

mutant background. (C) mEJP amplitudes are unchanged by loss of cmpy. (D)

Quantal content is reduced in cmpy mutants by 30% and is and rescued by

presynaptic expression of a wild-type Cmpy transgene in motorneurons. ***,

P<0.001. Recordings were conducted in 1.0mM Ca2+. n, number of muscles, one

muscle was recorded from per animal. Data is pooled from two independent

experiments, and represents the mean ± the SEM. (E-F) cmpy is necessary for

synaptic vesicle cycling. (E) Representative images of FM1-43 dye uptake by

NMJ4 in control and cmpy∆8 mutants after 90mM K+ depolarization. FM1-43

labels cmpy∆8 NMJs less efficiently than wild-type controls. Control corresponds

to heterozygous parental PBac elements, PBacf01736 / PBacf02482. Scale bar,

10μM. (F) Quantification of normalized FM1-43 labeling intensity at NMJ4. FM1-

43 labeling intensity is decreased by 30% at cmpy∆8 NMJs, mirroring the defect in

evoked synaptic transmission in (A). ***, P<0.001. n, number of NMJs from 18

animal. Data is pooled from three independent experiments, and represents the

mean ± SEM.

138

Figure 3.6

139

Figure 3.6 Continued

140

Figure 3.6: Model of Crimpy-dependent Gbb trafficking to the NMJ

Proposed models for Gbb trafficking to the NMJ. (A) Crimpy binds to precursor

Gbb early in the secretory pathway, sorting Gbb into DCVs of the regulated

secretory pathway in the trans-Golgi network. DCVs containing Crimpy and Gbb are trafficked to the NMJ, where they are docked and await activity-dependent cues that promote vesicle fusion with the plasma membrane and secretion of

DCV contents. Small vesicles of the constitutive secretory pathway are continuously secreted, independent of synaptic activity. (B) In cmpy mutants,

Gbb is mis-sorted into the constitutive secretory pathway, ultimately resulting in

NMJ overgrowth.

141

Table 3.1 Verification of transgene function at the NMJ.

142

Supplementary Figure 3.1

143

Supplementary Figure 3.1 (Continued)

144

Supplementary Figure 3.1: Cmpy-Venus localization at the NMJ

Representative confocal sections at NMJ4 in (A,B), and series projection in (C).

(A) A branch of synaptic boutons at NMJ4 of a UAScmpyVenus control body

wall. Anti-GFP does not label the NMJ of controls (Top). The presynaptic

compartment is labeled by the neuronal membrane marker anti-HRP (blue, top),

and the postsynaptic compartment is labeled by the subsynaptic reticulum

marker, anti-Dlg (pink, middle). (B) Cmpy-Venus puncta (green) partially

colocalize with the neuronal membrane at the cell surface (top). Cmpy-Venus

also localizes to an intracellular compartment within synaptic boutons. Cmpy-

Venus localization is restricted from the postsynaptic compartment labeled with

Dlg (middle). All three channels are merged in the bottom panel. (C) Cmpy-

Venus expression does not strongly overlap expression of the synaptic vesicle marker, DVGLUT. Scale bar: 10µm.

145

Supplementary Figure 3.2

146

Supplementary Figure 3.2: DVGLUT localization is increased at cmpy∆8

NMJs.

Representative confocal series projection of a terminal strand of type Ib synaptic

boutons at NMJ 6/7. (A) DVGLUT (anti-DVGLUT, green) labels synaptic boutons

diffusely in wild-type controls. Control corresponds to heterozygous parental

PBac elements, PBacf01736 / PBacf02482. The neuronal membrane is labeled with

anti-HRP in the rightmost panels (pink). (B) DVGLUT labeling intensity is

increased at cmpyΔ8 NMJs. (C) Quantification of the ratio of DVGLUT labeling

intensity relative to HRP labeling intensity for the given genotype. DVGLUT

intensity is increased 80% at cmpyΔ8 terminals. **, P<0.01. n, number of NMJs

scored from more than 10 animals. Scale bars: 10μm. Data is pooled from three independent experiments represents the mean ± SEM.

147

Supplementary Figure 3.3

148

Supplementary Figure 3.3: pMad accumulation is increased at cmpyΔ8

NMJs

Representative confocal series projection of a terminal strand of type Ib synaptic

boutons at NMJ 6/7. (A) pMad (anti-pMad, green) labels discrete puncta within

synaptic boutons in wild-type controls (top). Control corresponds to heterozygous

parental PBac elements, PBacf01736 / PBacf02482. The neuronal membrane is

labeled with anti-HRP in the rightmost panels (purple). pMad labeling intensity is

increased at cmpyΔ8 NMJs (bottom). (B) Quantification of the ratio of pMad

labeling intensity relative to HRP labeling intensity for the given genotype at the

NMJ. pMad accumulation is increased 2.7-fold at cmpyΔ8 terminals. (C)

Quantification of the ratio of pMad labeling intensity in motorneuron nuclei

relative to labeling intensity of the neuronal transcription factor, Elav. ***,

P<0.001. n, number of NMJs scored from more than 10 animals. Scale bars:

10μm. Data is pooled from three independent experiments represents the mean

± SEM.

149

CHAPTER 4: General Discussion

150

Challenging the classic BMP signaling paradigm at the Drosophila NMJ

Over a decade has passed since BMP signaling was unveiled as a critical regulator of both morphological expansion and synaptic transmission at the

Drosophila larval NMJ. Rescue studies suggested that muscle was the key source of the relevant BMP ligand Gbb (McCabe et al. 2003), establishing an unchallenged retrograde BMP signaling paradigm that has predominated the field since. Yet Gbb is not only expressed in the muscle, as one might predict if Gbb was exclusively a retrograde ligand that coordinates growth and function in motorneurons to match the muscle’s demand for synaptic input. Gbb is also expressed in the CNS (McCabe et al. 2003), where its secretion from postsynaptic motorneurons strengthens synaptic transmission between motorneurons and the presynaptic interneurons innervating them (Baines 2004).

Central expression of Gbb seems paradoxical for a presumably retrograde ligand, since presynaptic expression of Gbb would dilute the muscle-derived pool of Gbb ligand at the NMJ—the amount of which is thought to instruct presynaptic growth—thus uncoupling growth of the motorneuron terminal from that of the muscle.

In addition to identifying muscle as a potential source of growth-promoting ligand at the NMJ, rescues studies indicated that presynaptic Gbb regulates synaptic transmission. Initial rescue experiments using a Gbb transgene with leaky expression in the absence of a Gal4 driver in a gbb hypomorphic background suggested that motorneuronal expression of Gbb partially rescued defects in

151

synaptic transmission at gbb deficient NMJs, whereas panneuronal expression of

Gbb completely restored function (McCabe et al. 2003). In this study, postsynaptic expression of the leaky transgene partially restored function.

However a later study in a true gbb null background using a non-leaky transgene

revealed that postsynaptic expression of Gbb is incapable of conferring even

partial rescue of the defects in synaptic transmission at gbb null NMJs (Goold

and Davis 2007). Importantly, the later study revealed a continuous need for

BMP signaling at the NMJ to maintain proper baseline neurotransmission and

homeostatic synaptic plasticity independent of NMJ growth, as inhibiting BMP

signaling late during larval development after growth is completed impaired

synaptic function with no effect on morphology (Goold and Davis 2007). This

strongly supports the hypothesis that pro-neurotransmission BMP signaling in

motorneurons is initiated from a BMP pool distinct from the pro-growth signaling

pool; however the source of the synaptic transmission-promoting ligand is

uncertain. Indeed, an important insight from these studies has been largely

overlooked in the field: what is the function of motorneuronal Gbb at the NMJ?

Motorneuronal Gbb could be excluded from the NMJ, thus conferring the

proposed retrograde directionality of BMP signaling in the periphery. In line with

this, my initial studies identified Crimpy as an inhibitor of autocrine, pro-growth

BMP signaling in motorneurons. But, taking advantage of a novel HA-tagged Gbb

transgene that undergoes appropriate processing and drives overexpression

phenotypes comparable to wild-type transgenes, I found that Gbb is not excluded

152

from the motorneuron terminal. In fact Cmpy is necessary and sufficient for Gbb localization at the NMJ, supporting the idea that motorneuronal Gbb is functional in the periphery.

Considering the identified roles for neuronal Gbb in regulating synaptic transmission, it is probable that motorneuronal Gbb specifically regulates synaptic transmission at the NMJ. We knocked down Gbb in motorneurons by

RNA-interference and found that motorneuronal Gbb is normally dispensable for growth. Still, NMJ undergrowth in gbb mutants is rescued by expression of Gbb

in either muscle or motorneurons (McCabe et al. 2003, Goold and Davis 2007),

and we found that overexpression of wild-type Gbb in motorneurons is sufficient to drive overgrowth. Therefore, motorneuronal Gbb is capable of pro-growth signaling, while muscle-derived Gbb is incapable of rescuing defects in synaptic transmission in a gbb null background (Goold and Davis 2007). This challenges the established view of muscle-derived retrograde BMP signaling the periphery,

and calls for further clarification of the sources of active ligand that influence NMJ

development.

Activity-dependent secretion of motorneuronal Gbb and neurotransmission

Our findings that (1) Gbb-HA localization at presynaptic terminals depends on

Cmpy, (2) Cmpy and Gbb colocalize with a dense core vesicle marker within

motor nerves and at the NMJ, and (3) Gbb-HA levels are decreased at the NMJ

in response to synaptic activity, strongly indicate that Cmpy promotes Gbb

153

secretion from the NMJ in response to synaptic activity. Since disruption of

regulated secretory cargo sorting into DCVs diverts normally regulated cargo into

the constitutive secretory pathway (Burgess and Kelly 1987), we propose that

overgrowth at cmpy mutant terminals is due to constitutive secretion of Gbb at

the NMJ. However we cannot yet rule out the possibility that activity induces degradation of Gbb at the NMJ as a homeostatic response to sufficient synaptic input onto the postsynaptic muscle. Others have inhibited the ubiquitin proteasome system (UPS) with the pharmacological agents lactacystin and epoxomicin to assess levels of UPS-regulated proteins at the Drosophila NMJ

(Aravamudan and Broadie 2003, Speese et al. 2003). Assaying levels of Gbb-HA at the NMJ following stimulation with high K+ or blue light-activated channelrhodopsin in the presence of proteasome inhibitors should reveal if levels of Gbb-HA at the NMJ are sensitive to activity of the proteasome. We anticipate that if Gbb-HA is indeed secreted in response to activity, inhibiting the proteasome while acutely driving synaptic activity will have no effect on the decrease in Gbb-HA levels observed in response to activity.

Also, examining extracellular Gbb-HA levels following high K+ or

channelrhodopsin-mediated stimulation will help clarify our understanding of Gbb

secretion at the NMJ. We have been unable to detect secreted Gbb, likely due to

the tight control of extracellular BMP levels (Umulis et al. 2009). Still, optimization

of extracellular staining protocols should reveal an increase in Gbb-HA levels

outside of the neuronal membrane that mirrors decreased intracellular Gbb-HA

154

levels following stimulation. Together these experiments will tell if Gbb is indeed

released from presynaptic motorneuron terminals in response to synaptic activity.

As mentioned previously, early rescue experiments clearly demonstrate that

neuronal Gbb is necessary for proper basal synaptic transmission (McCabe et al.

2003, Goold and Davis 2007), while muscle-derived Gbb is dispensable for neurotransmission (Goold and Davis 2007). Our finding that activity regulates

Gbb levels at the NMJ is reminiscent of activity-dependent synaptic plasticity mediated by neurotrophin signaling in vertebrates. Neurotrophin secretion in response to activity strengthens synaptic transmission at vertebrate synapses by enhancing presynaptic release (Haubensak et al. 1998, Poo 2001), and TGFβs have recently been shown to undergo activity-dependent secretion (Lacmann et al. 2007) and promote presynaptic neurotransmitter release (Krieglstein et al.

2011). We therefore reasoned that activity-dependent secretion of Gbb at the

NMJ strengthens synaptic transmission. As activity-dependent regulation of Gbb levels at the NMJ is abolished in the absence of cmpy, rescue experiments evaluating synaptic transmission in a gbb cmpy double mutant background will provide insight into the role of activity-dependent Gbb secretion in modulating neurotransmission at the NMJ. We predict that, consistent with early rescue studies, expression of Gbb in motorneurons will rescue synaptic transmission in a gbb mutant background (McCabe et al. 2003), but that rescue of synaptic function will be decreased in gbb cmpy double mutants. Furthermore, knocking down Gbb expression in motorneurons specifically by RNA-interference in an

155

otherwise wild-type background and examining synaptic transmission will

complement rescue studies. If activity-dependent secretion of Gbb strengthens synaptic transmission, we envision Gbb knockdown in motorneurons will impair evoked synaptic transmission at the NMJ.

Evaluating synaptic transmission following overexpression of both Gbb and

Cmpy in motorneurons may reveal that levels of activity-dependent Gbb secretion strengthen synaptic transmission by enhancing presynaptic neurotransmitter release at the NMJ. We showed that overexpression of Gbb alone drives synaptic overgrowth, but co-overexpression of Gbb and Cmpy does not perturb growth at the NMJ. However this does not rule out the possibility that synaptic transmission is enhanced at NMJs co-overexpressing Gbb and Cmpy.

In fact, co-overexpression of Cmpy-Venus enhances Gbb-HA localization at the

NMJ. It will be fascinating to learn if evoked synaptic transmission is increased by co-overexpression of Gbb and Cmpy.

It is interesting to note that compromised BMP signaling in wit and gbb mutants results in detachment of the presynaptic membrane and floating electron dense

T-bodies that cluster clear-cored synaptic vesicles within the cytoplasm of synaptic boutons at the NMJ (Aberle et al. 2002, Marques et al. 2002, McCabe et al. 2003). These phenotypes could be considered ultrastructural hallmarks of synapse destabilization. As mentioned in the Introduction of this thesis, a non- canonical BMP signal transduction cascade involving activation of DLIMK-1

156

downstream of Wit promotes synapse stabilization at the NMJ, and presynaptic

expression of DLIMK-1 in a wit mutant background rescues defects in synaptic

transmission at wit mutant NMJs (Eaton and Davis 2005). An attractive

hypothesis for transduction of activity-dependent Gbb signaling in motorneurons

is that activation of DLIMK-1 downstream of Wit following activity-dependent

secretion of Gbb at the NMJ stabilizes active synapses.

In line with activity-dependent Gbb signaling regulating synaptic transmission

through a non-canonical pathway, we found that pMad levels at the NMJ of cmpy

mutants are increased 2.7-fold and nuclear pMad levels are unchanged. Yet synaptic transmission at cmpy mutant NMJs is impaired. This supports the idea that canonical BMP pathway activation alone is insufficient for proper synaptic transmission, while capable of driving NMJ overgrowth through a divergent pMad-mediated local mechanism. Several studies have highlighted the importance of maintaining a balance between nuclear and local pMad in regulating synaptic development (Higashi-Kovtun et al. 2010, McCabe et al.

2003, Merino et al. 2009), however the mechanism by which non-nuclear pMad affects NMJ growth remains open for investigation.

If activity-dependent Gbb signal transduction at the NMJ regulates synaptic transmission through activation of DLIMK-1, investigating genetic interactions between cmpy and DLIMK-1 may reveal synergistic defects in synaptic transmission. In the absence of proper sorting into the RSP, proteins that

157

normally undergo regulated secretion are diverted into the default constitutive

secretory pathway (Vazquez-Martinez et al. 2012). It is possible that residual

DLIMK-1 activation by constitutively-secreted motorneuronal Gbb preserves some pro-neurotransmission BMP signaling activity at the NMJ in the absence of cmpy. If this were the case, synaptic transmission might be more greatly impaired by reducing the gene dosage of DLIMK-1 in cmpy mutant background.

To further understand potential links between activity-dependent Gbb signaling at the NMJ and synapse stabilization, ultrastructural analysis of cmpyΔ8 NMJs would

be informative. As mentioned, gbb and wit mutant NMJs exhibit defects in

synapse stability, including presynaptic membrane detachment and floating T-

bodies. Our finding that DVGLUT and synaptotagmin levels are increased at

cmpyΔ8 NMJs calls to mind synaptic vesicle sequestering, floating T-bodies within

the cytoplasm of wit and gbb mutant synaptic boutons. Quantification of T-bodies

and clear-cored synaptic vesicles within synaptic boutons at cmpy mutant NMJs compared to controls would illuminate synapse stability in cmpy mutants and could provide a link between impaired neurotransmission and increased

DVGLUT and synaptotagmin levels at cmpyΔ8 NMJs. One could imagine that evoked synaptic transmission is impaired in cmpy mutants because free floating

T-bodies reduce the number of neurotransmitter-containing synaptic vesicles at cmpy mutant terminals available to achieve appropriate quantal release upon stimulation.

158

Cmpy is a novel sorting receptor for BMP delivery into the RSP

While considerable progress has been made in understanding the activity- dependent transcription and secretion of neurotrophic growth factors, namely

BDNF (Haubensak et al. 1998, Greenberg et al. 2009), very little is known about activity-dependent transcription and secretion of TGFβ peptides. TGFβs were only recently shown to undergo regulated secretion from neurons in response to activity (Lacmann et al. 2007). Furthermore, while BDNF is widely known to

regulate synaptic plasticity (Lu 2003), TGFβ ligands, and BMPs in particular, are

only recently emerging as regulators of synaptic transmission in addition to their

classic role as neuroprotective factors (Krieglstein et al. 2011). Given the

parallels between these pathways, it is likely that similar mechanisms are in

place to oversee TGFβ sorting into the RSP within vertebrate neurons. In

hippocampal neurons in vitro, the membrane protein Sortilin interacts with the

prodomain of BDNF, directing proBDNF into the RSP for activity-dependent

secretion (Chen et al. 2005). In cortical neurons in vitro, membrane-associated

Carboxypeptidase E (CPE) interacts with the mature domain of BDNF prior to

BDNF processing, similarly directing proBDNF into the RSP for activity-

dependent secretion (Lou et al. 2005).

We report here the first identification of a BMP sorting receptor in any species.

Like Sortilin binding to BDNF through its prodomain (Chen et al. 2005), we find

that Cmpy binds to the prodomain of Gbb. BDNF colocalizes with secretory

granule markers in both hippocampal and cortical neurons (Chen et al. 2005, Lou

159

et al. 2005). We similarly observed colocalization of both Gbb and Cmpy with a

marker for dense core secretory vesicles in Drosophila motor nerves and at the

NMJ. Gbb localization at the NMJ depends upon the presence of Cmpy. Since

devoted molecular chaperones in the ER actively sort and concentrate secretory

cargo into vesicles destined for the Golgi complex and further posttranslational

processing and sorting (Vazquez-Martinez et al. 2012), it is unlikely that colocalization of Gbb and Cmpy with dense core vesicle markers is an artifact of protein overexpression. Especially since Gbb protein levels at the NMJ are regulated by synaptic activity, and activity-dependent regulation of Gbb levels at the NMJ depends upon Cmpy. Together these studies demonstrate that Cmpy sorts Gbb into the RSP, however further ultrastructural characterization is needed to definitively place Gbb and Cmpy in DCVs as opposed to small, clear- cored SVs of the RSP. We propose that activity-dependent secretion of Gbb distinguishes motorneuronal Gbb from the muscle-derived pool of Gbb at the

NMJ, conferring its role in synaptic transmission as opposed to NMJ growth (Fig.

4.1). Yet the mechanisms downstream of activity-dependent Gbb release to

differentiate the two BMP signal transduction cascades within the motorneuron

have yet to be explored. It will be interesting to learn if non-canonical signaling

through DLIMK downstream of Wit transduces the motorneuronal BMP pathway.

We predict that cell-type specific sorting receptors actively direct BMP and TGFβ

ligands into the RSP in vertebrate neurons. Notably, we found that Cmpy binds to

precursor Gbb through an arginine-lysine (R-K) rich region, similar to interactions

160

between vertebrate DAN family member, Gremlin, and BMP4 (Sun et al. 2006).

This raises the exciting possibility that interactions mediated by R-K domains

may regulate BMP signaling by sorting BMPs into appropriate secretory

pathways. The cysteine-rich repeats (CRRs) of classic BMP antagonists like

Chordin, on the other hand, may inhibit or promote BMP signaling depending on

the cellular context. Cmpy contains a single degenerate CRR, suggesting that

protein sorting and direct regulation of BMP action, for example by inhibiting BMP

secretion or sequestering BMPs from receptors in the extracellular milieu, may

not be mutually exclusive events. It will be interesting to see if R-K rich domain

containing proteins like Gremlin direct BMP sorting into the RSP in vertebrate

neurons.

Mutations that disrupt neurotrophin-mediated synaptic plasticity are linked to

several neurological disorders in humans. For example, a single nucleotide

polymorphism in BDNF that disrupts its interaction with Sortilin and subsequent

delivery into the RSP is associated with memory impairment, Parkinson’s

disease, depression, and bipolar disorder (Momose et al. 2002, Neves-Pereira et al. 2002, Hariri et al. 2003, Sen et al. 2003). The identification and

characterization of additional growth factor sorting receptors could have

implications in human disease pathology. Moreover, developing a better

understanding of how key developmental signaling pathways diverge within a single cell type could provide a means to target one branch of signaling while permitting another branch of signaling to remain intact. For example, constitutive

161

secretion of growth factors like BMPs could confer neuroprotection within a given

neuronal population, while regulated secretion could modulate

neurotransmission. If synaptic transmission was adversely enhanced for some

reason, it might be beneficial to inhibit the regulated pool of ligand within that

neuronal population with a small molecule antagonist, while leaving the

neuroprotective form of signaling intact. Understanding intracellular signaling

pathway branch points like the sorting of growth factors into the RSP versus the

CSP provides a molecular target to selectively inhibit one aspect of signaling without affecting all aspects of signaling, and could have huge implications for developing novel small molecules to treat human disease.

162

Figure 4.1

163

Figure 4.1: Proposed Model for Distinguishing BMP Pathways at the NMJ

At least two potential sources of BMP ligand regulate development of the NMJ

through distinct pathways: Gbb derived from the muscle (green), and Gbb

derived from the motorneuron (purple). Cmpy (orange) sorts motorneuronal Gbb

into DCVs (blue circle) of the regulated secretory pathway, and activity-

dependent Gbb secretion from motorneurons promotes synaptic transmission at

the NMJ. The mechanism differentiating motorneuronal BMP signal transduction

from muscle-derived BMP signal transduction is unknown. An appealing

hypothesis is that motorneuron-derived Gbb signals through a non-canonical

BMP pathway by binding the type II BMP receptor Wit (blue) and activating

DLIMK (yellow) to regulate synaptic transmission. Release of Gbb from postsynaptic muscle, presumably through constitutive secretory vesicles (red circle) promotes growth of the presynaptic terminal through canonical BMP signal transduction involving Wit (blue), type I receptors Sax/Tkv (red), the R-Smad

Mad (tan), and the Co-Smad Medea (grey).

164

Aberle, H., A. P. Haghighi, R. D. Fetter, B. D. McCabe, T. R. Magalhaes and C. S. Goodman (2002). "wishful thinking encodes a BMP type II receptor that regulates synaptic growth in Drosophila." Neuron 33(4): 545-558.

Affolter, M. and K. Basler (2007). "The Decapentaplegic morphogen gradient: from pattern formation to growth regulation." Nat Rev Genet 8(9): 663-674.

Akiyama, T., G. Marques and K. A. Wharton (2012). "A large bioactive BMP ligand with distinct signaling properties is produced by alternative proconvertase processing." Sci Signal 5(218): ra28.

Alder, J., K. J. Lee, T. M. Jessell and M. E. Hatten (1999). "Generation of cerebellar granule neurons in vivo by transplantation of BMP-treated neural progenitor cells." Nat Neurosci 2(6): 535-540.

Allan, D. W., S. E. St Pierre, I. Miguel-Aliaga and S. Thor (2003). "Specification of neuropeptide cell identity by the integration of retrograde BMP signaling and a combinatorial transcription factor code." Cell 113(1): 73-86.

Ambrosio, A. L., V. F. Taelman, H. X. Lee, C. A. Metzinger, C. Coffinier and E. M. De Robertis (2008). "Crossveinless-2 Is a BMP feedback inhibitor that binds Chordin/BMP to regulate Xenopus embryonic patterning." Dev Cell 15(2): 248-260.

Aravamudan, B. and K. Broadie (2003). "Synaptic Drosophila UNC-13 is regulated by antagonistic G-protein pathways via a proteasome-dependent degradation mechanism." J Neurobiol 54(3): 417-438.

Arora, K., M. B. O'Connor and R. Warrior (1996). "BMP signaling in Drosophila embryogenesis." Ann N Y Acad Sci 785: 80-97.

Ataman, B., J. Ashley, M. Gorczyca, P. Ramachandran, W. Fouquet, S. J. Sigrist and V. Budnik (2008). "Rapid activity-dependent modifications in synaptic structure and function require bidirectional Wnt signaling." Neuron 57(5): 705-718.

Attisano, L. and J. L. Wrana (2002). "Signal transduction by the TGF-beta superfamily." Science 296(5573): 1646-1647.

Atwood, H. L., C. K. Govind and C. F. Wu (1993). "Differential ultrastructure of synaptic terminals on ventral longitudinal abdominal muscles in Drosophila larvae." J Neurobiol 24(8): 1008-1024.

Bachiller, D., J. Klingensmith, C. Kemp, J. A. Belo, R. M. Anderson, S. R. May, J. A. McMahon, A. P. McMahon, R. M. Harland, J. Rossant and E. M. De Robertis (2000). "The organizer factors Chordin and Noggin are required for mouse forebrain development." Nature 403(6770): 658-661.

Baines, R. A. (2004). "Synaptic strengthening mediated by bone morphogenetic protein- dependent retrograde signaling in the Drosophila CNS." J Neurosci 24(31): 6904-6911.

Ballard, S. L., J. Jarolimova and K. A. Wharton (2010). "Gbb/BMP signaling is required to maintain energy homeostasis in Drosophila." Dev Biol 337(2): 375-385.

165

Barlowe, C. K. and E. A. Miller (2013). "Secretory protein biogenesis and traffic in the early secretory pathway." Genetics 193(2): 383-410.

Bekinschtein, P., M. Cammarota, I. Izquierdo and J. H. Medina (2008). "BDNF and memory formation and storage." Neuroscientist 14(2): 147-156.

Bier, E. (2008). "Intriguing extracellular regulation of BMP signaling." Dev Cell 15(2): 176-177.

Brand, A. H. and N. Perrimon (1993). "Targeted gene expression as a means of altering cell fates and generating dominant phenotypes." Development 118(2): 401-415.

Broadie, K. and M. Bate (1993). "Activity-dependent development of the neuromuscular synapse during Drosophila embryogenesis." Neuron 11(4): 607-619.

Broihier, H. T. and J. B. Skeath (2002). "Drosophila homeodomain protein dHb9 directs neuronal fate via crossrepressive and cell-nonautonomous mechanisms." Neuron 35(1): 39-50.

Burgess, T. L. and R. B. Kelly (1987). "Constitutive and regulated secretion of proteins." Annu Rev Cell Biol 3: 243-293.

Chen, Z. Y., A. Ieraci, H. Teng, H. Dall, C. X. Meng, D. G. Herrera, A. Nykjaer, B. L. Hempstead and F. S. Lee (2005). "Sortilin controls intracellular sorting of brain-derived neurotrophic factor to the regulated secretory pathway." J Neurosci 25(26): 6156-6166.

Chesnutt, C., L. W. Burrus, A. M. Brown and L. Niswander (2004). "Coordinate regulation of neural tube patterning and proliferation by TGFbeta and WNT activity." Dev Biol 274(2): 334-347.

Collins, C. A. and A. DiAntonio (2007). "Synaptic development: insights from Drosophila." Curr Opin Neurobiol 17(1): 35-42.

Conley, C. A., R. Silburn, M. A. Singer, A. Ralston, D. Rohwer-Nutter, D. J. Olson, W. Gelbart and S. S. Blair (2000). "Crossveinless 2 contains cysteine-rich domains and is required for high levels of BMP-like activity during the formation of the cross veins in Drosophila." Development 127(18): 3947-3959.

Constam, D. B. and E. J. Robertson (1999). "Regulation of bone morphogenetic protein activity by pro domains and proprotein convertases." J Cell Biol 144(1): 139-149.

Coyle, I. P., Y. H. Koh, W. C. Lee, J. Slind, T. Fergestad, J. T. Littleton and B. Ganetzky (2004). "Nervous wreck, an SH3 adaptor protein that interacts with Wsp, regulates synaptic growth in Drosophila." Neuron 41(4): 521-534.

Cui, Y., F. Jean, G. Thomas and J. L. Christian (1998). "BMP-4 is proteolytically activated by furin and/or PC6 during vertebrate embryonic development." Embo J 17(16): 4735-4743.

166

Daniels, R. W., C. A. Collins, K. Chen, M. V. Gelfand, D. E. Featherstone and A. DiAntonio (2006). "A single vesicular glutamate transporter is sufficient to fill a synaptic vesicle." Neuron 49(1): 11-16.

Daniels, R. W., C. A. Collins, M. V. Gelfand, J. Dant, E. S. Brooks, D. E. Krantz and A. DiAntonio (2004). "Increased expression of the Drosophila vesicular glutamate transporter leads to excess glutamate release and a compensatory decrease in quantal content." J Neurosci 24(46): 10466-10474.

De Robertis, E. M. (2009). "Spemann's organizer and the self-regulation of embryonic fields." Mech Dev 126(11-12): 925-941.

Decotto, E. and E. L. Ferguson (2001). "A positive role for Short gastrulation in modulating BMP signaling during dorsoventral patterning in the Drosophila embryo." Development 128(19): 3831-3841.

Degnin, C., F. Jean, G. Thomas and J. L. Christian (2004). "Cleavages within the prodomain direct intracellular trafficking and degradation of mature bone morphogenetic protein-4." Mol Biol Cell 15(11): 5012-5020.

Dickman, D. K., Z. Lu, I. A. Meinertzhagen and T. L. Schwarz (2006). "Altered synaptic development and active zone spacing in endocytosis mutants." Curr Biol 16(6): 591-598.

Dieni, S., T. Matsumoto, M. Dekkers, S. Rauskolb, M. S. Ionescu, R. Deogracias, E. D. Gundelfinger, M. Kojima, S. Nestel, M. Frotscher and Y. A. Barde (2012). "BDNF and its pro-peptide are stored in presynaptic dense core vesicles in brain neurons." J Cell Biol 196(6): 775-788.

Dudu, V., T. Bittig, E. Entchev, A. Kicheva, F. Julicher and M. Gonzalez-Gaitan (2006). "Postsynaptic mad signaling at the Drosophila neuromuscular junction." Curr Biol 16(7): 625-635.

Eaton, B. A. and G. W. Davis (2005). "LIM Kinase1 controls synaptic stability downstream of the type II BMP receptor." Neuron 47(5): 695-708.

Ellis, J. E., L. Parker, J. Cho and K. Arora (2010). "Activin signaling functions upstream of Gbb to regulate synaptic growth at the Drosophila neuromuscular junction." Dev Biol 342(2): 121-133.

Featherstone, D. E. and K. Broadie (2000). "Surprises from Drosophila: genetic mechanisms of synaptic development and plasticity." Brain Res Bull 53(5): 501-511.

Fouquet, W., D. Owald, C. Wichmann, S. Mertel, H. Depner, M. Dyba, S. Hallermann, R. J. Kittel, S. Eimer and S. J. Sigrist (2009). "Maturation of active zone assembly by Drosophila Bruchpilot." J Cell Biol 186(1): 129-145.

Francois, V., M. Solloway, J. W. O'Neill, J. Emery and E. Bier (1994). "Dorsal-ventral patterning of the Drosophila embryo depends on a putative negative growth factor encoded by the short gastrulation gene." Genes Dev 8(21): 2602-2616.

167

Frank, C. A., M. J. Kennedy, C. P. Goold, K. W. Marek and G. W. Davis (2006). "Mechanisms underlying the rapid induction and sustained expression of synaptic homeostasis." Neuron 52(4): 663-677.

Frasch, M. (1995). "Induction of visceral and cardiac mesoderm by ectodermal Dpp in the early Drosophila embryo." Nature 374(6521): 464-467.

Fung, W. Y., K. F. Fat, C. K. Eng and C. K. Lau (2007). "crm-1 facilitates BMP signaling to control body size in Caenorhabditis elegans." Dev Biol 311(1): 95-105.

Garcia Abreu, J., C. Coffinier, J. Larrain, M. Oelgeschlager and E. M. De Robertis (2002). "Chordin-like CR domains and the regulation of evolutionarily conserved extracellular signaling systems." Gene 287(1-2): 39-47.

Gazzerro, E. and E. Canalis (2006). "Bone morphogenetic proteins and their antagonists." Rev Endocr Metab Disord 7(1-2): 51-65.

Goold, C. P. and G. W. Davis (2007). "The BMP ligand Gbb gates the expression of synaptic homeostasis independent of synaptic growth control." Neuron 56(1): 109-123.

Greenberg, M. E., B. Xu, B. Lu and B. L. Hempstead (2009). "New insights in the biology of BDNF synthesis and release: implications in CNS function." J Neurosci 29(41): 12764- 12767.

Hariri, A. R., T. E. Goldberg, V. S. Mattay, B. S. Kolachana, J. H. Callicott, M. F. Egan and D. R. Weinberger (2003). "Brain-derived neurotrophic factor val66met polymorphism affects human memory-related hippocampal activity and predicts memory performance." J Neurosci 23(17): 6690-6694.

Haubensak, W., F. Narz, R. Heumann and V. Lessmann (1998). "BDNF-GFP containing secretory granules are localized in the vicinity of synaptic junctions of cultured cortical neurons." J Cell Sci 111 ( Pt 11): 1483-1493.

Higashi-Kovtun, M. E., T. J. Mosca, D. K. Dickman, I. A. Meinertzhagen and T. L. Schwarz (2010). "Importin-beta11 regulates synaptic phosphorylated mothers against decapentaplegic, and thereby influences synaptic development and function at the Drosophila neuromuscular junction." J Neurosci 30(15): 5253-5268.

James, R. E. and H. T. Broihier (2011). "Crimpy inhibits the BMP homolog Gbb in motoneurons to enable proper growth control at the Drosophila neuromuscular junction." Development 138(15): 3273-3286.

Jin, Y. (2002). "Synaptogenesis: insights from worm and fly." Curr Opin Neurobiol 12(1): 71-79.

Johansen, J., M. E. Halpern, K. M. Johansen and H. Keshishian (1989). "Stereotypic morphology of glutamatergic synapses on identified muscle cells of Drosophila larvae." J Neurosci 9(2): 710-725.

Kawabata, M., T. Imamura and K. Miyazono (1998). "Signal transduction by bone morphogenetic proteins." Cytokine Growth Factor Rev 9(1): 49-61.

168

Keshishian, H., K. Broadie, A. Chiba and M. Bate (1996). "The drosophila neuromuscular junction: a model system for studying synaptic development and function." Annu Rev Neurosci 19: 545-575.

Keshishian, H., A. Chiba, T. N. Chang, M. S. Halfon, E. W. Harkins, J. Jarecki, L. Wang, M. Anderson, S. Cash, M. E. Halpern and et al. (1993). "Cellular mechanisms governing synaptic development in Drosophila melanogaster." J Neurobiol 24(6): 757-787.

Keshishian, H. and Y. S. Kim (2004). "Orchestrating development and function: retrograde BMP signaling in the Drosophila nervous system." Trends Neurosci 27(3): 143-147.

Khalsa, O., J. W. Yoon, S. Torres-Schumann and K. A. Wharton (1998). "TGF-beta/BMP superfamily members, Gbb-60A and Dpp, cooperate to provide pattern information and establish cell identity in the Drosophila wing." Development 125(14): 2723-2734.

Kim, S., Y. P. Wairkar, R. W. Daniels and A. DiAntonio (2010). "The novel endosomal membrane protein Ema interacts with the class C Vps-HOPS complex to promote endosomal maturation." J Cell Biol 188(5): 717-734.

Kinna, G., G. Kolle, A. Carter, B. Key, G. J. Lieschke, A. Perkins and M. H. Little (2006). "Knockdown of zebrafish crim1 results in a bent tail phenotype with defects in somite and vascular development." Mech Dev 123(4): 277-287.

Kittel, R. J., C. Wichmann, T. M. Rasse, W. Fouquet, M. Schmidt, A. Schmid, D. A. Wagh, C. Pawlu, R. R. Kellner, K. I. Willig, S. W. Hell, E. Buchner, M. Heckmann and S. J. Sigrist (2006). "Bruchpilot promotes active zone assembly, Ca2+ channel clustering, and vesicle release." Science 312(5776): 1051-1054.

Kolle, G., K. Georgas, G. P. Holmes, M. H. Little and T. Yamada (2000). "CRIM1, a novel gene encoding a cysteine-rich repeat protein, is developmentally regulated and implicated in vertebrate CNS development and organogenesis." Mechanisms of Development 90(2): 181-193.

Kolle, G., A. Jansen, T. Yamada and M. Little (2003). "In ovo electroporation of Crim1 in the developing chick spinal cord." Dev Dyn 226(1): 107-111.

Krause, C., O. Korchynskyi, K. de Rooij, S. E. Weidauer, D. J. de Gorter, R. L. van Bezooijen, S. Hatsell, A. N. Economides, T. D. Mueller, C. W. Lowik and P. ten Dijke (2010). "Distinct modes of inhibition by sclerostin on bone morphogenetic protein and Wnt signaling pathways." J Biol Chem 285(53): 41614-41626.

Krieglstein, K., F. Zheng, K. Unsicker and C. Alzheimer (2011). "More than being protective: functional roles for TGF-beta/activin signaling pathways at central synapses." Trends Neurosci 34(8): 421-429.

Kunnapuu, J., I. Bjorkgren and O. Shimmi (2009). "The Drosophila DPP signal is produced by cleavage of its proprotein at evolutionary diversified furin-recognition sites." Proc Natl Acad Sci U S A 106(21): 8501-8506.

169

Kuromi, H. and Y. Kidokoro (2005). "Exocytosis and endocytosis of synaptic vesicles and functional roles of vesicle pools: lessons from the Drosophila neuromuscular junction." Neuroscientist 11(2): 138-147.

Lacmann, A., D. Hess, G. Gohla, E. Roussa and K. Krieglstein (2007). "Activity- dependent release of transforming growth factor-beta in a neuronal network in vitro." Neuroscience 150(3): 647-657.

Lahey, T., M. Gorczyca, X. X. Jia and V. Budnik (1994). "The Drosophila tumor suppressor gene dlg is required for normal synaptic bouton structure." Neuron 13(4): 823-835.

Landgraf, M., S. Roy, A. Prokop, K. VijayRaghavan and M. Bate (1999). "even-skipped determines the dorsal growth of motor axons in Drosophila." Neuron 22(1): 43-52.

Landgraf, M. and S. Thor (2006). "Development of Drosophila motoneurons: specification and morphology." Semin Cell Dev Biol 17(1): 3-11.

Le Roux, P., S. Behar, D. Higgins and M. Charette (1999). "OP-1 enhances dendritic growth from cerebral cortical neurons in vitro." Exp Neurol 160(1): 151-163.

Lein, P., M. Johnson, X. Guo, D. Rueger and D. Higgins (1995). "Osteogenic protein-1 induces dendritic growth in rat sympathetic neurons." Neuron 15(3): 597-605.

Levitan, E. (2012). H. Broihier.

Liu, A. and L. A. Niswander (2005). "Bone morphogenetic protein signalling and vertebrate nervous system development." Nat Rev Neurosci 6(12): 945-954.

Lou, H., S. K. Kim, E. Zaitsev, C. R. Snell, B. Lu and Y. P. Loh (2005). "Sorting and activity-dependent secretion of BDNF require interaction of a specific motif with the sorting receptor carboxypeptidase e." Neuron 45(2): 245-255.

Lu, B. (2003). "BDNF and activity-dependent synaptic modulation." Learn Mem 10(2): 86-98.

Ma, X. and T. Xie (2011). "Stem cells: keeping BMP signaling local." Curr Biol 21(19): R809-811.

Marques, G. (2005). "Morphogens and synaptogenesis in Drosophila." J Neurobiol 64(4): 417-434.

Marques, G., H. Bao, T. E. Haerry, M. J. Shimell, P. Duchek, B. Zhang and M. B. O'Connor (2002). "The Drosophila BMP type II receptor Wishful Thinking regulates neuromuscular synapse morphology and function." Neuron 33(4): 529-543.

Marrus, S. B., S. L. Portman, M. J. Allen, K. G. Moffat and A. DiAntonio (2004). "Differential localization of glutamate receptor subunits at the Drosophila neuromuscular junction." J Neurosci 24(6): 1406-1415.

170

Mathew, D., B. Ataman, J. Chen, Y. Zhang, S. Cumberledge and V. Budnik (2005). "Wingless signaling at synapses is through cleavage and nuclear import of receptor DFrizzled2." Science 310(5752): 1344-1347.

McCabe, B. D., S. Hom, H. Aberle, R. D. Fetter, G. Marques, T. E. Haerry, H. Wan, M. B. O'Connor, C. S. Goodman and A. P. Haghighi (2004). "Highwire regulates presynaptic BMP signaling essential for synaptic growth." Neuron 41(6): 891-905.

McCabe, B. D., G. Marques, A. P. Haghighi, R. D. Fetter, M. L. Crotty, T. E. Haerry, C. S. Goodman and M. B. O'Connor (2003). "The BMP homolog Gbb provides a retrograde signal that regulates synaptic growth at the Drosophila neuromuscular junction." Neuron 39(2): 241-254.

Merino, C., J. Penney, M. Gonzalez, K. Tsurudome, M. Moujahidine, M. B. O'Connor, E. M. Verheyen and P. Haghighi (2009). "Nemo kinase interacts with Mad to coordinate synaptic growth at the Drosophila neuromuscular junction." J Cell Biol 185(4): 713-725.

Miech, C., H. U. Pauer, X. He and T. L. Schwarz (2008). "Presynaptic local signaling by a canonical wingless pathway regulates development of the Drosophila neuromuscular junction." J Neurosci 28(43): 10875-10884.

Miller, C. M., A. Page-McCaw and H. T. Broihier (2008). "Matrix metalloproteinases promote motor axon fasciculation in the Drosophila embryo." Development 135(1): 95- 109.

Mishina, Y. (2003). "Function of bone morphogenetic protein signaling during mouse development." Front Biosci 8: d855-869.

Miyazono, K., Y. Kamiya and M. Morikawa (2010). "Bone morphogenetic protein receptors and signal transduction." J Biochem 147(1): 35-51.

Mizutani, C. M., Q. Nie, F. Y. Wan, Y. T. Zhang, P. Vilmos, R. Sousa-Neves, E. Bier, J. L. Marsh and A. D. Lander (2005). "Formation of the BMP activity gradient in the Drosophila embryo." Dev Cell 8(6): 915-924.

Momose, Y., M. Murata, K. Kobayashi, M. Tachikawa, Y. Nakabayashi, I. Kanazawa and T. Toda (2002). "Association studies of multiple candidate genes for Parkinson's disease using single nucleotide polymorphisms." Ann Neurol 51(1): 133-136.

Nahm, M., S. Kim, S. K. Paik, M. Lee, S. Lee, Z. H. Lee, J. Kim, D. Lee, Y. C. Bae and S. Lee (2010). "dCIP4 (Drosophila Cdc42-interacting protein 4) restrains synaptic growth by inhibiting the secretion of the retrograde Glass bottom boat signal." J Neurosci 30(24): 8138-8150.

Nahm, M., A. A. Long, S. K. Paik, S. Kim, Y. C. Bae, K. Broadie and S. Lee (2010). "The Cdc42-selective GAP rich regulates postsynaptic development and retrograde BMP transsynaptic signaling." J Cell Biol 191(3): 661-675.

Nechipurenko, I. V. and H. T. Broihier (2012). "FoxO limits microtubule stability and is itself negatively regulated by microtubule disruption." J Cell Biol 196(3): 345-362.

171

Neves-Pereira, M., E. Mundo, P. Muglia, N. King, F. Macciardi and J. L. Kennedy (2002). "The brain-derived neurotrophic factor gene confers susceptibility to bipolar disorder: evidence from a family-based association study." Am J Hum Genet 71(3): 651-655.

O'Connor-Giles, K., L. Ho and B. Ganetzky (2008). "Nervous Wreck Interacts with Thickveins and the Endocytic Machinery to Attenuate Retrograde BMP Signaling during Synaptic Growth." Neuron 58(4): 507-518.

O'Connor, M. B., D. Umulis, H. G. Othmer and S. S. Blair (2006). "Shaping BMP morphogen gradients in the Drosophila embryo and pupal wing." Development 133(2): 183-193.

Odden, J. P., S. Holbrook and C. Q. Doe (2002). "Drosophila HB9 is expressed in a subset of motoneurons and interneurons, where it regulates gene expression and axon pathfinding." J Neurosci 22(21): 9143-9149.

Packard, M., E. S. Koo, M. Gorczyca, J. Sharpe, S. Cumberledge and V. Budnik (2002). "The Drosophila Wnt, wingless, provides an essential signal for pre- and postsynaptic differentiation." Cell 111(3): 319-330.

Packard, M., D. Mathew and V. Budnik (2003). "Wnts and TGF beta in synaptogenesis: old friends signalling at new places." Nat Rev Neurosci 4(2): 113-120.

Parks, A. L., K. R. Cook, M. Belvin, N. A. Dompe, R. Fawcett, K. Huppert, L. R. Tan, C. G. Winter, K. P. Bogart, J. E. Deal, M. E. Deal-Herr, D. Grant, M. Marcinko, W. Y. Miyazaki, S. Robertson, K. J. Shaw, M. Tabios, V. Vysotskaia, L. Zhao, R. S. Andrade, K. A. Edgar, E. Howie, K. Killpack, B. Milash, A. Norton, D. Thao, K. Whittaker, M. A. Winner, L. Friedman, J. Margolis, M. A. Singer, C. Kopczynski, D. Curtis, T. C. Kaufman, G. D. Plowman, G. Duyk and H. L. Francis-Lang (2004). "Systematic generation of high- resolution deletion coverage of the Drosophila melanogaster genome." Nature Genetics 36(3): 288-292.

Pennisi, D. J., L. Wilkinson, G. Kolle, M. L. Sohaskey, K. Gillinder, M. J. Piper, J. W. McAvoy, F. J. Lovicu and M. H. Little (2007). "Crim1KST264/KST264 mice display a disruption of the Crim1 gene resulting in perinatal lethality with defects in multiple organ systems." Dev Dyn 236(2): 502-511.

Petersen, S. A., R. D. Fetter, J. N. Noordermeer, C. S. Goodman and A. DiAntonio (1997). "Genetic analysis of glutamate receptors in Drosophila reveals a retrograde signal regulating presynaptic transmitter release." Neuron 19(6): 1237-1248.

Piccolo, S., E. Agius, B. Lu, S. Goodman, L. Dale and E. M. De Robertis (1997). "Cleavage of Chordin by Xolloid metalloprotease suggests a role for proteolytic processing in the regulation of Spemann organizer activity." Cell 91(3): 407-416.

Poo, M. M. (2001). "Neurotrophins as synaptic modulators." Nat Rev Neurosci 2(1): 24- 32.

Raftery, L. A. and D. J. Sutherland (1999). "TGF-beta family signal transduction in Drosophila development: from Mad to Smads." Dev Biol 210(2): 251-268.

172

Ralston, A. and S. S. Blair (2005). "Long-range Dpp signaling is regulated to restrict BMP signaling to a crossvein competent zone." Dev Biol 280(1): 187-200.

Rao, S., C. Lang, E. S. Levitan and D. L. Deitcher (2001). "Visualization of neuropeptide expression, transport, and exocytosis in Drosophila melanogaster." J Neurobiol 49(3): 159-172.

Rawson, J. M., M. Lee, E. L. Kennedy and S. B. Selleck (2003). "Drosophila neuromuscular synapse assembly and function require the TGF-beta type I receptor saxophone and the transcription factor Mad." J Neurobiol 55(2): 134-150.

Ray, R. P. and K. A. Wharton (2001). "Context-dependent relationships between the BMPs gbb and dpp during development of the Drosophila wing imaginal disk." Development 128(20): 3913-3925.

Ross, J. J., O. Shimmi, P. Vilmos, A. Petryk, H. Kim, K. Gaudenz, S. Hermanson, S. C. Ekker, M. B. O'Connor and J. L. Marsh (2001). "Twisted gastrulation is a conserved extracellular BMP antagonist." Nature 410(6827): 479-483.

Rossignol, P., S. Collier, M. Bush, P. Shaw and J. H. Doonan (2007). "Arabidopsis POT1A interacts with TERT-V(I8), an N-terminal splicing variant of telomerase." J Cell Sci 120(Pt 20): 3678-3687.

Saitoe, M., T. L. Schwarz, J. A. Umbach, C. B. Gundersen and Y. Kidokoro (2001). "Absence of junctional glutamate receptor clusters in Drosophila mutants lacking spontaneous transmitter release." Science 293(5529): 514-517.

Sanchez-Soriano, N., G. Tear, P. Whitington and A. Prokop (2007). "Drosophila as a genetic and cellular model for studies on axonal growth." Neural Dev 2: 9.

Sanyal, S. (2009). "Genomic mapping and expression patterns of C380, OK6 and D42 enhancer trap lines in the larval nervous system of Drosophila." Gene Expr Patterns 9(5): 371-380.

Schinder, A. F. and M. Poo (2000). "The neurotrophin hypothesis for synaptic plasticity." Trends Neurosci 23(12): 639-645.

Schroll, C., T. Riemensperger, D. Bucher, J. Ehmer, T. Voller, K. Erbguth, B. Gerber, T. Hendel, G. Nagel, E. Buchner and A. Fiala (2006). "Light-induced activation of distinct modulatory neurons triggers appetitive or aversive learning in Drosophila larvae." Curr Biol 16(17): 1741-1747.

Schuster, C. M., G. W. Davis, R. D. Fetter and C. S. Goodman (1996). "Genetic dissection of structural and functional components of synaptic plasticity. I. Fasciclin II controls synaptic stabilization and growth." Neuron 17(4): 641-654.

Schuster, C. M., G. W. Davis, R. D. Fetter and C. S. Goodman (1996). "Genetic dissection of structural and functional components of synaptic plasticity. II. Fasciclin II controls presynaptic structural plasticity." Neuron 17(4): 655-667.

173

Sen, S., R. M. Nesse, S. F. Stoltenberg, S. Li, L. Gleiberman, A. Chakravarti, A. B. Weder and M. Burmeister (2003). "A BDNF coding variant is associated with the NEO personality inventory domain neuroticism, a risk factor for depression." Neuropsychopharmacology 28(2): 397-401.

Serpe, M., A. Ralston, S. S. Blair and M. B. O'Connor (2005). "Matching catalytic activity to developmental function: tolloid-related processes Sog in order to help specify the posterior crossvein in the Drosophila wing." Development 132(11): 2645-2656.

Serpe, M., D. Umulis, A. Ralston, J. Chen, D. J. Olson, A. Avanesov, H. Othmer, M. B. O'Connor and S. S. Blair (2008). "The BMP-binding protein Crossveinless 2 is a short- range, concentration-dependent, biphasic modulator of BMP signaling in Drosophila." Dev Cell 14(6): 940-953.

Shi, Y. and J. Massague (2003). "Mechanisms of TGF-beta signaling from cell membrane to the nucleus." Cell 113(6): 685-700.

Shimmi, O., A. Ralston, S. S. Blair and M. B. O'Connor (2005). "The crossveinless gene encodes a new member of the Twisted gastrulation family of BMP-binding proteins which, with Short gastrulation, promotes BMP signaling in the crossveins of the Drosophila wing." Dev Biol 282(1): 70-83.

Shimmi, O., D. Umulis, H. Othmer and M. B. O'Connor (2005). "Facilitated transport of a Dpp/Scw heterodimer by Sog/Tsg leads to robust patterning of the Drosophila blastoderm embryo." Cell 120(6): 873-886.

Sossin, W. S., J. M. Fisher and R. H. Scheller (1990). "Sorting within the regulated secretory pathway occurs in the trans-Golgi network." J Cell Biol 110(1): 1-12.

Specht, H., H. Peterziel, M. Bajohrs, H. H. Gerdes, K. Krieglstein and K. Unsicker (2003). "Transforming growth factor beta2 is released from PC12 cells via the regulated pathway of secretion." Mol Cell Neurosci 22(1): 75-86.

Speese, S. D., N. Trotta, C. K. Rodesch, B. Aravamudan and K. Broadie (2003). "The ubiquitin proteasome system acutely regulates presynaptic protein turnover and synaptic efficacy." Curr Biol 13(11): 899-910.

Sun, J., F. F. Zhuang, J. E. Mullersman, H. Chen, E. J. Robertson, D. Warburton, Y. H. Liu and W. Shi (2006). "BMP4 activation and secretion are negatively regulated by an intracellular gremlin-BMP4 interaction." J Biol Chem 281(39): 29349-29356.

Sun, M., M. J. Thomas, R. Herder, M. L. Bofenkamp, S. B. Selleck and M. B. O'Connor (2007). "Presynaptic contributions of chordin to hippocampal plasticity and spatial learning." J Neurosci 27(29): 7740-7750.

Sweeney, S. T. and G. W. Davis (2002). "Unrestricted synaptic growth in spinster-a late endosomal protein implicated in TGF-beta-mediated synaptic growth regulation." Neuron 36(3): 403-416.

Tessier-Lavigne, M. and C. S. Goodman (1996). "The molecular biology of axon guidance." Science 274(5290): 1123-1133.

174

Thibault, S. T., M. A. Singer, W. Y. Miyazaki, B. Milash, N. A. Dompe, C. M. Singh, R. Buchholz, M. Demsky, R. Fawcett, H. L. Francis-Lang, L. Ryner, L. M. Cheung, A. Chong, C. Erickson, W. W. Fisher, K. Greer, S. R. Hartouni, E. Howie, L. Jakkula, D. Joo, K. Killpack, A. Laufer, J. Mazzotta, R. D. Smith, L. M. Stevens, C. Stuber, L. R. Tan, R. Ventura, A. Woo, I. Zakrajsek, L. Zhao, F. Chen, C. Swimmer, C. Kopczynski, G. Duyk, M. L. Winberg and J. Margolis (2004). "A complementary transposon tool kit for Drosophila melanogaster using P and piggyBac." Nat Genet 36(3): 283-287.

Thoenen, H. (1995). "Neurotrophins and neuronal plasticity." Science 270(5236): 593- 598.

Thomas, K. and A. Davies (2005). "Neurotrophins: a ticket to ride for BDNF." Curr Biol 15(7): R262-264.

Tomizawa, K., H. Matsui, E. Kondo, K. Miyamoto, M. Tokuda, T. Itano, S. Nagahata, T. Akagi and O. Hatase (1995). "Developmental alteration and neuron-specific expression of bone morphogenetic protein-6 (BMP-6) mRNA in rodent brain." Brain Res Mol Brain Res 28(1): 122-128.

Topol, L. Z., M. Marx, D. Laugier, N. N. Bogdanova, N. V. Boubnov, P. A. Clausen, G. Calothy and D. G. Blair (1997). "Identification of drm, a novel gene whose expression is suppressed in transformed cells and which can inhibit growth of normal but not transformed cells in culture." Mol Cell Biol 17(8): 4801-4810.

Tripodi, M. and S. Arber (2012). "Regulation of motor circuit assembly by spatial and temporal mechanisms." Curr Opin Neurobiol 22(4): 615-623.

Turrigiano, G. G. (1999). "Homeostatic plasticity in neuronal networks: the more things change, the more they stay the same." Trends Neurosci 22(5): 221-227.

Umulis, D., M. B. O'Connor and S. S. Blair (2009). "The extracellular regulation of bone morphogenetic protein signaling." Development 136(22): 3715-3728.

Vactor, D. V., H. Sink, D. Fambrough, R. Tsoo and C. S. Goodman (1993). "Genes that control neuromuscular specificity in Drosophila." Cell 73(6): 1137-1153.

Vazquez-Martinez, R., A. Diaz-Ruiz, F. Almabouada, Y. Rabanal-Ruiz, F. Gracia- Navarro and M. M. Malagon (2012). "Revisiting the regulated secretory pathway: from frogs to human." Gen Comp Endocrinol 175(1): 1-9.

Wagh, D. A., T. M. Rasse, E. Asan, A. Hofbauer, I. Schwenkert, H. Durrbeck, S. Buchner, M. C. Dabauvalle, M. Schmidt, G. Qin, C. Wichmann, R. Kittel, S. J. Sigrist and E. Buchner (2006). "Bruchpilot, a protein with homology to ELKS/CAST, is required for structural integrity and function of synaptic active zones in Drosophila." Neuron 49(6): 833-844.

Walsh, D. W., C. Godson, D. P. Brazil and F. Martin (2010). "Extracellular BMP- antagonist regulation in development and disease: tied up in knots." Trends Cell Biol 20(5): 244-256.

175

Wang, X., R. E. Harris, L. J. Bayston and H. L. Ashe (2008). "Type IV collagens regulate BMP signalling in Drosophila." Nature 455(7209): 72-77.

Wang, X., W. R. Shaw, H. T. Tsang, E. Reid and C. J. O'Kane (2007). "Drosophila spichthyin inhibits BMP signaling and regulates synaptic growth and axonal microtubules." Nat Neurosci 10(2): 177-185.

Weng, Y. L., N. Liu, A. DiAntonio and H. T. Broihier (2011). "The cytoplasmic adaptor protein Caskin mediates Lar signal transduction during Drosophila motor axon guidance." J Neurosci 31(12): 4421-4433.

Wharton, K. A., J. M. Cook, S. Torres-Schumann, K. de Castro, E. Borod and D. A. Phillips (1999). "Genetic analysis of the bone morphogenetic protein-related gene, gbb, identifies multiple requirements during Drosophila development." Genetics 152(2): 629- 640.

Wilkinson, L., G. Kolle, D. Wen, M. Piper, J. Scott and M. Little (2003). "CRIM1 regulates the rate of processing and delivery of bone morphogenetic proteins to the cell surface." J Biol Chem 278(36): 34181-34188.

Withers, G. S., D. Higgins, M. Charette and G. Banker (2000). "Bone morphogenetic protein-7 enhances dendritic growth and receptivity to innervation in cultured hippocampal neurons." Eur J Neurosci 12(1): 106-116.

Zakin, L. and E. M. De Robertis (2010). "Extracellular regulation of BMP signaling." Curr Biol 20(3): R89-92.

Zhang, X., L. Bao and G. Q. Ma (2010). "Sorting of neuropeptides and neuropeptide receptors into secretory pathways." Prog Neurobiol 90(2): 276-283.

176