EXAMINING DYNAMIC ASPECTS OF PRESYNAPTIC TERMINAL FORMATION

VIA

LIVE CONFOCAL MICROSCOPY

By

LUKE ANDREW DASCENZO BURY

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Dissertation Adviser: Dr. Shasta Sabo

Department of Pharmacology

CASE WESTERN RESERVE UNIVERSITY

August, 2015

II

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of

Luke A. D. Bury

candidate for the Doctor of Philosophy degree*.

Committee Chair

Ruth Siegel

Committee Member

Shasta Sabo

Committee Member

Lynn Landmesser

Committee Member

Brian McDermott

Committee Member

Edwin Levitan

Date of Defense

June 7th

, 2015

*We also certify that written approval has been obtained

For any proprietary material contained therein

III

Table of Contents

List of Tables VII

List of Figures VIII

Abstract X

Chapter 1: General Introduction 1

Basic synaptic physiology 2

Presynaptic distribution and composition 3

STVs and PTVs – transport vesicles for efficient presynaptic protein delivery 4

STV and PTV recruitment into development synapses – a dynamic process 12

Synaptogenic signaling through trans-synaptic adhesion proteins 14

Chapter 2: Coordinated trafficking of synaptic vesicle and active zone proteins prior

to synapse formation 24

Abstract 25

Introduction 26

Results 29

STVs and PTVs can move together 29

STVs and PTVs display similar pausing characteristics 33

IV

STVs and PTVs pause simultaneously at the same sites in the axon 36

Recruitment of a PTV enhances accumulation of additional PTVs

at sites of synapse formation 44

STV pausing is only mildly dependent on PTVs 46

Materials and Methods 51

Acknowledgements 55

Chapter 3: Dynamic mechanisms of neuroligin-dependent presynaptic terminal

assembly in living cortical neurons 56

Abstract 57

Introduction 58

Results 60

Synaptic protein recruitment fluctuates frequently and rapidly during

assembly of individual presynaptic terminals 60

Levels of synaptic vesicle proteins are highly labile at developing

axo-dendritic contacts 69

Synaptophysin is preferentially stabilized at clustered neurexin 72

Actin polymerization is occasionally co-localized with neurexin

clustering 76

Levels of recruited synaptophysin and neurexin are strongly

correlated at developing presynaptic terminals 77

V

Trans-synaptic signaling recruits synaptic proteins without

substantially altering their transport 80

Synaptic vesicle and active zone protein transport vesicles are not

attracted to sites of trans-synaptic signaling 85

Initial arrival of STVs and PTVs to a developing synapse is likely

uncoordinated 91

Materials and Methods 93

Acknowledgements 98

Chapter 4: Imaging presynaptic dynamics during mouse cortical development via

two-photon confocal microscopy 99

Introduction 100

Results 101

Materials and Methods 111

Chapter 5: Discussion 115

Is STV and PTV transport coordinated? 115

What causes STV and PTV pausing? 117

Why don’t all vesicles pause at any given site? 119

A model of coordinated transport 120

VI

What are the dynamics of presynaptic protein recruitment at developing

terminals? 122

Presynaptic proteins are trapped at, but not attracted to, developing synapse 122

Synaptic proteins are labile at developing synapses 124

A model of neuroligin-neurexin mediated presynaptic protein recruitment 125

A model for dynamic presynaptic protein recruitment to a developing

synapse 127

In vivo synaptic dynamics 128

Future outlooks – directions for future investigations 130

Conclusion 134

Bibliography 135

VII

List of Tables

Table 1: STV and PTV proteins 6

Table 2: Trans-synaptic adhesion partners 17

VIII

List of Figures

Figure 1.1: Newly synthesized presynaptic proteins are packaged into STVs and PTVs

and then transported along axonal microtubules for delivery to developing

synapses 7

Figure 1.2: Mechanisms of regulation of STV and PTV transport 11

Figure 1.3: Recruitment of presynaptic cargo downstream of neurexin dependent

synaptogenic adhesion 21

Figure 2.1: STVs and PTVs move together 31

Figure 2.2: PTV pausing is qualitatively and quantitatively similar to STV pausing 34

Figure 2.3: STVs and PTVs pause at the same sites 38

Figure 2.4: STVs and PTVs preferentially pause at the same sites at the same time 42

Figure 2.5: Multiple PTVs are attracted to the same sites 46

Figure 2.6: A direct interaction between STVs and PTVs cannot account for the

attraction of STVs to pause sites that contain PTVs 49

Figure 3.1: Levels of synaptic vesicle protein enrichment at individual trans-synaptic

adhesion sites are modulated throughout recruitment 63

Figure 3.2: Modulation of synaptic vesicle protein recruitment occurs through two

distinct mechanisms 67

Figure 3.3: Synaptic vesicle protein recruitment to axo-dendritic contacts is similar

that at induced trans-synaptic signaling sites 70

IX

Figure 3.4: Synaptic vesicle protein accumulation occurs at clustered neurexin

and correlates with levels of clustered neurexin 74

Figure 3.5: Neurexin and synaptophysin levels fluctuate over the course of minutes

at axo-dendritic contact sites and neuroligin induced developing synapses 79

Figure 3.6: Trans-synaptic signaling has little effect on synaptic vesicle or active

zone protein transport 83

Figure 3.7: Trans-synaptic signaling does not attract synaptic vesicle or active zone

proteins 89

Figure 4.1: Synaptophysin-tdtomato expression in transgenic knock-in mouse 103

Figure 4.2: Synaptophysin-tdtomato is localized to presynaptic terminals 105

Figure 4.3: Synaptophysin-tdtomato puncta both split and merge in vivo 108

Figure 4.4: Instantaneous velocity and mean displacement decrease as

development progresses 110

Figure 5.1: Models for coordination of recruitment of STVs and PTVs to the same

place at the same time 121

Figure 5.2: Model of synaptic vesicle protein transport vesicle (STV) recruitment

to a developing presynaptic terminal 128

X

Examining Dynamic Aspects of Presynaptic Terminal

Formation via Live Confocal Microscopy

Abstract

By

LUKE ANDREW DASCENZO BURY

To create a presynaptic terminal, molecular signaling events must be orchestrated within

a number of subcellular compartments. In the soma, presynaptic proteins need to be

synthesized, packaged together, and attached to microtubule motors for shipment through

the axon. Within the axon, transport of presynaptic packages is regulated in order to

ensure that developing synapses receive an adequate supply of components. At individual

axonal sites, extracellular interactions must be translated into intracellular signals that can

incorporate mobile transport vesicles into the nascent presynaptic terminal. Even once the

initial recruitment process is complete, the components and subsequent functionality of

presynaptic terminals need to constantly be remodeled. Perhaps most remarkably, all of

these processes need to be coordinated in space and time. In this dissertation, I will

discuss how these dynamic cellular processes occur in neurons of the central nervous

system in order to generate presynaptic terminals in the brain, and describe experiments

to further elucidate these mechanisms.

1

Chapter 1: General Introduction

Over 100 years ago, seminal work by Ramon y Cajal led to the development of

the neuron doctrine, which states that the nervous system is comprised of individual cells

(1, 2). Inspired by Cajal, Charles Sherrington proposed the existence of synapses that

would allow communication between these cells (2, 3). From these early beginnings, it

has become clear that the synaptic connections between neurons are indispensible for

proper nervous system function and for life itself.

Synaptogenesis is important throughout life. Extensive synapse formation during

neonatal and early postnatal development is responsible for initial establishment of the

neural circuitry in the brain. However, synapse formation does not cease upon

maturation. Synaptogenesis, in concert with synapse elimination, is a constant process

that occurs throughout life as circuits in the brain are modified in response to varying

external stimuli. The ability of the brain to constantly remodel its circuitry over time is

critical in order to successfully adapt to and live within changing environments over the

course of an organism’s life.

Studying how synapses form is necessary to understand both how the brain

functions normally and how this goes awry in disease. Aberrations in synaptogenesis

have been linked to a number of neurodevelopmental disorders, including autism and

schizophrenia (4-7). In addition, poor synapse formation has been implicated in a wide

variety of non-developmental neurological diseases, including addiction and depression

(8-11). Furthermore, synapse loss is one of the hallmarks of neurodegenerative diseases

such as Alzheimer’s disease, Parkinson’s disease, and Huntington’s disease (12-15).

While it is not clear whether synapse formation is decreased or synapse elimination is

2

enhanced in these diseases (or both), enhancing synaptogenesis is an important

therapeutic strategy to mitigate the effects of these devastating diseases (16).

Basic synaptic physiology

The basic purpose of a synapse in the brain is to enable transmission of electrical

signals from one neuron to another. There are two types of synapses in the brain that

accomplish this feat: electrical and chemical. Electrical synapses enable a direct

connection between two neurons through the formation of gap junctions (17, 18). Gap

junctions are specialized connections mediated by connexin proteins that result in a pore

forming between two connected cells (19, 20). This pore not only physically links the

cells, it also electrically links them by allowing rapid diffusion of ions from the

cytoplasm in either direction across the junction. Therefore, unlike chemical synapses,

electrical information can be transmitted by either cell to its connected partner. Because

of this near-instantaneous bi-directional signaling capability, electrical synapses are often

found between inhibitory neurons in the brain where they are utilized to synchronize

inhibitory signals (17, 21).

While critical for normal nervous system function, electrical synapses make up

only a small portion of synapses in the brain. Chemical synapses take the role as the

major synaptic type in both the peripheral and central nervous systems. At chemical

synapses, cells are physically separated by the synaptic cleft and electrical information is

transferred between neurons via neurotransmitters. As action potentials reach presynaptic

terminals, the resulting depolarization opens voltage-gated calcium channels within the

presynaptic plasma membrane, leading to an influx of calcium. This, in turn, causes

synaptic vesicles (SVs) loaded with neurotransmitter to fuse with the active zone

3

membrane, via exocytic SNARE proteins. As SVs fuse, neurotransmitters are released

into the synaptic cleft, which stimulates the postsynaptic cell through the opening of

neurotransmitter-gated ion channels and/or the activation of metabotropic

neurotransmitter receptors. For more thorough reviews of presynaptic function please see

the following (22-25).

The neurotransmitter and postsynaptic neurotransmitter receptors vary according

to the neuronal cell type and location. In the brain, glutamate is the main excitatory

neurotransmitter, while gamma-aminobutyric acid (GABA) functions as the main

inhibitory neurotransmitter. However, many other neurotransmitters are produced and

released in the brain, including dopamine, serotonin, acetylcholine, norepinephrine, ATP,

cannabinoids, and various peptides (26-33). These other neurotransmitters can function in

either an excitatory or inhibitory role, depending on the circuit they are utilized in and the

type of receptor they activate.

Presynaptic distribution and composition

The location and number of presynaptic terminals within an axon varies. For

motor neurons, presynaptic structures known as terminaux boutons are located at the

distal end of the projecting axon where it makes contact with the muscle. In contrast,

locally projecting neurons of the cerebral cortex typically form “en passant” presynaptic

terminals along much of the length of unmyelinated axons. Strikingly, the number of

presynaptic terminals per axon in the cortex can number in the thousands (34-36).

Recently, an elegant study showed that an “average” presynaptic terminal in the

cortex or cerebellum is composed of approximately 300,000 individual molecules (37).

4

Although the number and diversity of proteins at the synapse is striking, many

presynaptic proteins can be classified into two important groups: (i) synaptic vesicle (SV)

associated proteins and (ii) proteins of the cytomatrix at the active zone (CAZ) (38, 39).

SVs are 35-45nm spherical vesicles that fuse with the synaptic plasma membrane to

release neurotransmitter (25, 40-43). SV-associated proteins control neurotransmitter

uptake into SVs, trafficking of SVs within the synapse, and fusion of SVs with the

plasma membrane. CAZ proteins are structural components of the presynaptic terminal,

tethering SVs to the active zone and priming them for fusion with the plasma membrane.

In addition, SNARE proteins within the CAZ interact with the v-SNARE

synaptobrevin/VAMP2 on synaptic vesicles to enable SV fusion with the plasma

membrane (44).

STVs and PTVs – transport vesicles for efficient presynaptic protein delivery

Building presynaptic terminals in the CNS poses unique challenges for the

neuron. Most of the proteins that comprise the presynaptic side of the synapse are

translated in the soma. Since some axons extend for millimeters, presynaptic components

oftentimes need to be transported long distances from where they were made.

Furthermore, presynaptic terminals form along an individual axon in an unsynchronized

manner. This requires specific assembly of presynaptic proteins and structures at the right

place at the right time.

To facilitate efficient delivery to developing synapses, groups of presynaptic

components are sorted into precursor organelles (Figure 1.1). Specifically, SV proteins

are incorporated into Synaptic vesicle precursor protein Transport Vesicles (STVs), while

CAZ proteins are packaged into Piccolo-bassoon Transport Vesicles (PTVs) (45, 46).

5

STVs are tubulo-vesicular structures that are heterogeneous in size and shape (47-49).

These golgi-derived vesicles can cycle with the axonal membrane and primarily contain

proteins that are associated with SVs, including synapsin, synaptophysin, SV2, VGlut1,

VAMP2, Rab3a, synaptotagmin and amphiphysin (Table 1) (47-58). PTVs primarily

exist as 80nm dense-core vesicles (59, 60) that are derived from the trans-golgi network

(61, 62), and they carry proteins of the CAZ, including piccolo, bassoon, syntaxin, RIM,

Munc-18, ELKS2/CAST, SNAP-25, and n-cadherin (Table 1) (49, 59-62).

6

Table 1: STV and PTV proteins

Synaptic proteins Motors Motor Linkage

STVs

Synaptic vesicle proteins, including:

synapsin, synaptophysin, SV2, VGlut1,

VAMP2, Rab3a, synaptotagmin and

amphiphysin

kinesin-3/KIF1A

and KIF1B

syd-2/liprin-

DENN/MADD, via

Rab 3

SAM-4

PI(4,5)P(2)

kinesin-1/KIF5 ?

dynein dynactin complex

PTVs

Active zone cytomatrix proteins, including:

piccolo, bassoon, syntaxin, RIM, Munc-18,

ELKS2/CAST, SNAP-25, and n-cadherin

kinesin-1/KIF5 syntabulin, via

syntaxin

Dynein bassoon

7

Figure 1.1: Newly synthesized presynaptic proteins are packaged into STVs and

PTVs and then transported along axonal microtubules for delivery to developing

synapses. STVs and PTVs are derived from the trans-Golgi network. CAZ and SV-

associated proteins are sorted into PTVs and STVs, respectively. STVs and PTVs then

attach to kinesins and potentially dynein microtubule (MT) motors. Attachment can occur

through direct binding of motors and their cargo or facilitated via a linking protein. STVs

and PTVs are shipped through the axon on MTs. STV and PTV transport is coordinated,

and these organelles can move together. Axonal MTs are arranged with their minus end

toward the soma, suggesting that entry into the axon requires kinesin plus-end directed

motors.

8

Motors

After STVs and PTVs are formed, they are tethered to microtubule (MT) motors

and shipped through the axon (Figure 1.1). STVs and PTVs are associated with both

kinesin and dynein motors, which move toward and away from MT plus ends,

respectively (63). Since axonal MTs are oriented with their plus ends facing distally (64-

67), kinesins drive vesicles anterogradely. STVs are transported via the kinesin-3/KIF1A

and KIF1B motors and kinesin-1/KIF5, while PTVs are transported via kinesin-1/KIF5

(Table 1) (52, 68-79). Dynein drives both PTVs and STVs towards the soma (Table 1)

(80-86). Because STVs and PTVs are likely linked to anterograde and retrograde motors

simultaneously, they typically move in a bi-directional fashion (47, 48, 50, 53, 58, 60, 73,

87, 88). This movement is frequently interrupted by brief pauses. While the exact cause

of these pauses is unknown, the sites in the axon where STVs pause are preferential sites

of presynaptic terminal formation (51). Previously, it was not known whether STV and

PTV transport is coordinated within the axon. The experiments described here in Chapter

2 attempt to determine the coordination of STV and PTV transport in order to identify a

potential novel mechanism of rapid synapse formation via the instantaneous recruitment

of SV and CAZ components to a developing synapse.

Linkage to motors

Both PTVs and STVs can attach to kinesins through linker proteins, and

regulation of this interaction is critical for synapse development (Table 1). PTVs are

primarily linked to KIF5B through syntabulin (73, 74). Syntabulin simultaneously binds

to the KIF5B heavy chain and the CAZ protein syntaxin (74). Interrupting this interaction

disrupts PTV transport in the axon and leads to a decrease in the number of functional

9

presynaptic terminals (73). A number of kinesin/cargo linking proteins enable transport

of STVs. Invertebrate studies have identified syd-2/liprin-as a link between STVs and

kinesins (77, 89, 90). Disrupting this link leads to defective axonal transport and poor

incorporation of presynaptic components into synapses (77, 89). In addition, the protein

DENN/MADD (Differentially Expressed in Normal and Neoplastic cells/MAP kinase

Activating Death Domain) mediates transport of STVs in mammalian neurons by

simultaneously binding KIF1 and the STV-associated small GTPase Rab3 (91).

DENN/MADD only binds to the GTP-associated form of Rab3 and can also function as a

Rab3 GDP-GTP exchange factor (92-94). Consistent with a broader role for GTPases in

regulation of the STV/KIF1 linkage and subsequent synapse formation, the GTPase Arl-8

is critical for proper presynaptic localization in C. elegans (95). Specifically, the GTP-

bound form of Arl-8 directly interacts with UNC-104/KIF1A to ensure that aggregation

of presynaptic components does not occur prematurely in proximal axonal sections (95,

96). Further work in C. elegans has revealed that the SV associated protein SAM-4 also

interacts with the cargo-binding domain of unc-104/KIF1A and mediates STV transport

(97). In addition, unc-104/KIF1A can directly bind to STVs through

phosphatidylinositol-4,5-biphosphate (PI(4,5)P(2)), an integral lipid component of the

STV membrane, and disrupting this interaction leads to aberrant STV trafficking (98, 99).

The physical links between presynaptic transport vesicles and dynein are less well

characterized. Dynein consists of two large heavy chains that mediate ATP hydrolysis

and MT binding. The heavy chains are dimerized by five different intermediate/light

chains. These light chains also link dynein to adaptor complexes that mediate cargo

binding (84, 100). One of these adaptor complexes, dynactin, links STVs to dynein, as

10

alterations to its p150Glued

subunit leads to defective axonal transport of SV proteins (81,

85). In contrast, PTVs can directly interact with dynein light chains via bassoon, and

disrupting this interaction leads to aberrant trafficking of PTVs and altered protein

composition at synapses (83).

Additional mechanisms for regulation of STV and PTV transport

STV and PTV transport is regulated through a variety of dynamic mechanisms

(Figure 1.2). For example, phosphorylation of kinesin reduces binding to STVs (75, 101).

In addition, post-translational modification of MTs affects motor binding to MTs (102-

105). Moreover, STV and PTV movement through the axon can be altered by physical

impediments, such as reaching the end of a MT, binding of MT-associated proteins

(MAPs), or a traffic jam caused by additional motors and cargo binding the same MT

(106, 107). Neuronal activity can also affect axonal transport, possibly through calcium

influx within the axon (51, 108). The number and types of motors on individual vesicles

can vary as well, affecting transport (109). Finally, actin-based transport via myosins may

contribute to trafficking of presynaptic components, especially within the axonal initial

segment and at sites of synaptic protein recruitment (51, 103, 110-113).

11

Figure 1.2: Mechanisms of regulation of STV and PTV transport. Trafficking of

STVs and PTVs within the axon can be modulated in a number of ways. Regulation of

the following direct interactions between motor, cargo, and/or cytoskeletal elements can

potentially alter STV and/or PTV transport: (1) kinesin attachment to MTs, (2) kinesin

phosphorylation, (3) linker molecules between kinesin and cargo, (4) binding of myosin

motors to cargo, (5) binding of myosin motors to actin and/or actin polymerization at

specific sites in the axon, (6) direct physical interaction of kinesin or dynein with cargo,

(7) dynein attachment to MTs. Indirect regulation involving a number of other factors

might also affect STV and PTV trafficking: (1) motors and cargo reaching the ends of

MTs, and/or switching to adjacent MTs, (2) MT-associated proteins (MAPs) that alter

transport, (3) increased cytosolic Ca2+

, via influx and/or intracellular sources, (4) traffic

jams with other motors and cargo (synaptic or non-synaptic) on the same MT, (5) post-

translational modifications of MTs, such as acetylation or de-tyrosination. It is not yet

clear which of these mechanisms are employed to deposit synaptic material at sites of

synapse formation.

12

For a presynaptic terminal to form, STVs and PTVs must be deposited at sites of

axo-dendritic contact, requiring local cessation of transport. While many of the

mechanisms described above regulate transport and are important for synapse formation,

it remains unclear which of these mechanisms are used locally at sites of presynaptic

terminal formation. Some of these mechanisms may contribute to synapse assembly by

globally regulating transport within the axon, essentially controlling the supply of

synaptic building materials. For instance, in C. elegans, disrupting axonal transport of

presynaptic components leads to aberrant synapse formation (95, 96), and KIF1A is

critical for synaptogenesis in the hippocampus (114). Conversely, other mechanisms may

be engaged specifically at sites of synapse assembly in order to stop transport and deposit

the building materials. It will be important to distinguish between these possibilities in

the future. This will likely require local manipulation of each of these mechanisms at

sites of presynaptic assembly combined with high-resolution time-lapse imaging.

STV and PTV recruitment into developing synapses – a dynamic process

Initial recruitment of presynaptic proteins into a developing synapse can be rapid.

PTVs and STVs can be recruited to novel synaptogenic contacts within minutes (53, 115-

117). STV and PTV co-recruitment suggests that a functional synapse can form within

this time scale (118, 119). Indeed, both STVs and PTVs can instantly become trapped at

sites of synaptogenic signaling as they are trafficked through the axon (88, 115, 116).

Because STVs can cycle with the plasma membrane and likely release neurotransmitter

prior to recruitment (47, 48, 50, 51, 54-58, 120), new sites of recruitment probably have

at least some immediate functional capability. However, STV cycling is mechanistically

and functionally different than SV cycling at mature synapses (54, 56). The time required

13

for functional maturation is unclear. Live imaging of developing synapses followed by

retrospective electron microscopy indicated that the time required to accumulate a full

complement of synaptic structures is variable and often takes hours (121, 122). In

addition, it can take hours for axonal regions exposed to a postsynaptic synaptogenic

signal to become saturated with newly formed synapses (88, 115).

Questions also remain about the order and potential coordination of STV and PTV

recruitment. Since many CAZ proteins tether SVs at mature synapses, PTVs are typically

thought to arrive at developing synapses first, where they can then recruit STVs (123). In

support of this theory, the average time it takes for PTVs to be recruited to developing

synapses tends to be shorter than the average time for STV recruitment, albeit within

minutes of each other (116, 118). However, initial recruitment of STVs and PTVs has

only been observed in separate cells (i.e. observing STV recruitment in one neuron and

PTV recruitment in a different neuron). In both of the above studies, the time-course of

recruitment for SV and CAZ proteins overlaps (116, 118). In addition, STVs and PTVs

can co-localize with each other within the axon, suggesting that an entire complement of

synaptic proteins might be recruited to a developing synapse simultaneously (49, 87).

Furthermore, clustering of the trans-synaptic adhesion protein neurexin (see below)

within the axon leads to the formation of presynaptic terminals (124, 125). Finally, live

imaging of STVs and PTVs in separate axons of zebrafish embryos, followed by

retrospective immunocytochemistry identified STVs as the first transport vesicle that

arrives at the synapse (126). Therefore, live time-lapse imaging of both STVs and PTVs

within the same axon is needed to fully understand this critical event in the formation of a

presynaptic terminal.

14

Sites of Preferential Synapse Formation

Synapses form at discrete sites within neurons and are restricted in their overall

size (41, 127). Even when synaptogenic signaling is initiated in a continuous axonal area

tens of m in length, presynaptic terminals appear at only certain points within this

region (51, 88, 115, 128, 129). In addition, sites in the axon where STVs and PTVs pause

are sites of preferential synapse development (51, 87). Pausing at these preferential sites

might act as a mechanism for localizing presynaptic proteins to these areas, allowing

rapid incorporation of STVs and PTVs into the developing synapse because these

components are already present at the site once synaptogenic signaling commences.

However, this hypothesis needs to be tested directly. Many dendritic filopodia make

transient contact with neighboring axons, with only a portion of axo-dendritic contacts

becoming stabilized (130-135). Therefore, dendritic filopodia may sample the axon to

find the suitable synaptic sites.

The molecular mechanisms responsible for creating preferred sites of synapse

formation within the axon are unknown. One possibility is that axonal synaptogenic

adhesion proteins are involved. For example, distinct patterns or states of neurexin

oligomerization might identify preferred sites of synapse formation. However, the

interaction of neurexin with neuroligin is likely not required for initial formation of these

sites, as they can be identified prior to axo-dendritic contact (51). Future studies directed

at uncovering the molecular components responsible for these preferential sites will be

critical to understand the initial processes behind presynaptic terminal formation.

Synaptogenic signaling through trans-synaptic adhesion proteins

15

For a presynaptic terminal to form, STVs and PTVs must be recruited to a

specific site in the axon in response to a synaptogenic signal. This process is triggered by

physical contact between two cells – in the brain, typically between an axon and dendrite.

When axo-dendritic contact is established, binding occurs in trans between the

extracellular domains of adhesion proteins that are localized to axons and dendrites. This,

in turn, leads to synaptogenic signaling within the axon and dendrite, subsequent

recruitment of synaptic components to both sides of the contact, and eventual formation

of a functional synapse (136).

A number of synaptogenic trans-synaptic adhesion interactions have been

identified (Table 2). Differential expression of these proteins between neuronal

populations may help organize brain circuitry by controlling the specificity of synapse

formation (136-138). For example, two isoforms of neuroligin - neuroligin-1 and

neuroligin-2 - are specifically localized to excitatory and inhibitory synapses,

respectively (139, 140). This specificity appears to be driven by the axon, arising from

the interaction preferences of presynaptic neurexins found in excitatory and inhibitory

axons (141, 142). Likewise, isoforms of SynCAM are differentially expressed in regions

of the brain and display specific binding patterns between isoforms (143, 144). Binding

specificity can be mediated by alternative splicing. The neuroligin-neurexin interaction is

extensively regulated through this mechanism, and different binding pairs exhibit

different functional properties, including differences in the rates of presynaptic terminal

formation (115, 145-152).

Trans-synaptic adhesion proteins also display redundant functions during synapse

formation. Mice lacking neuroligin-1, 2, and 3 have a normal number of synapses (153).

16

Likewise, synapse number is normal in mice lacking all isoforms of the -neurexin gene

(154). Although synaptic function is altered in these models, the ability to form synapses

remains intact. However, the relative expression of neuroligin between neighboring cells

in the brain does regulate synapse number (155). This suggests that the

neuroligin/neurexin interaction is important for synapse development in a competitive,

non-cell-autonomous manner. In contrast, increasing or knocking out SynCAM increases

or decreases hippocampal synapse number, respectively (156). While alterations in

synapse number point to a synaptogenic function of SynCAM, the magnitudes of the

changes were approximately 10-20% (156), illustrating that other synaptogenic signals

are involved in this process. In fact, there is no individual molecule that is absolutely

critical for synapse formation in the brain. Therefore, determining how the multitude of

trans-synaptic adhesion signals cooperate to control synapse formation will be critical for

a full understanding of synapse and circuit formation in the brain.

17

Table 2: Trans-synaptic adhesion partners

Presynaptic Postsynaptic References

Homophilic

SynCAM

SynCAM (129, 156, 157)

N-Cadherin

N-Cadherin (158-161)

Amyloid Precursor Protein

(APP)

Amyloid precursor protein

(APP)

(162)

Heterophilic

Neurexin Neuroligin (124, 128, 139, 163, 164)

Neurexin Leucine Rich Repeat

Transmembrane Protein 2

(LRRTM2)

(165, 166)

Neurexin

GluR2 + Cerebellin 1

Precursor Protein

(167)

Protein Tyrosine

Phosphatase Receptor T

(PRPRT)

Neuroligin (168)

Protein Tyrosine

Phosphatase Sigma (PTP)

TrkC (169)

Protein Tyrosine

Phosphatase Delta (PTP)

Interleukin-1-receptor

accessory protein like 1

(IL1RAP)

(170-172)

Ephrin-B EphB (173, 174)

18

Latrophilin 1

Lasso/Teneurin-2 (175)

Netrin-G1, Netrin-G2,

Leukocyte Common

Antigen-Related Protein

(LAR)

Netrin-G Ligands (NGL-

1,2,3)

(176-178)

19

Signaling downstream of trans-synaptic adhesion

The downstream signaling pathways generated through trans-synaptic adhesion

are less clear, especially presynaptically. The interaction between neurexin and neuroligin

is the best characterized of the trans-synaptic proteins (Figure 1.3), and many of the

presynaptic signaling mechanisms that act down-stream of neurexin are likely to be

involved in synapse assembly induced by other trans-synaptic interactions.

The MAGUK protein CASK can bind to neurexin (179). CASK, in turn, binds to

syndecan-2 and protein 4.1 (180, 181). Synaptic proteins might be recruited to this

complex either through actin-mediated recruitment via filamentous-actin (F-actin)

nucleation on protein 4.1 or through phosphorylation of various substrates by CASK

itself (182, 183). Mint1 also binds to neurexin and can link neurexin, CASK, Veli/MALS,

and the active zone protein Munc-18 into a single complex (184, 185). This

CASK/Mint1/Veli complex may also play a role in SynCAM and APP mediated

presynaptic development (129, 162, 186). Furthermore, the tandem-PDZ protein syntenin

can directly interact with neurexin and is also linked to the CASK/Mint2/Veli complex

through syndecan-2 (187, 188). Syntenin also interacts with the active zone protein

CAST1/ERC2/ELKS via its PDZ domain, which in turn, binds RIM1, liprin-, bassoon,

and piccolo (188-192). Piccolo facilitates F-actin formation, which can lead to

recruitment of additional synaptic components (193), while bassoon interacts directly

with dynein to regulate recruitment of CAZ components (83).

Liprin- interacts with a number of these complexes, and disrupting this

interaction leads to aberrant synapse formation (194). Liprin- (syd-2 in C. elegans) is

20

involved in active zone morphology and recruitment of active zone proteins (195-201).

Interestingly, mutations that inhibit liprin- binding to the CASK/Mint1/Veli complex

have been linked to X-linked mental retardation (202). Because liprin- also links STVs

to KIF1, these interactions might be critical for initial recruitment of STVs into the

synapse (77, 89, 97, 203). Liprin-α acts downstream of syd-1 during presynaptic

development in C. elegans (196, 197). In Drosophila, Syd-1 binds to neurexin through its

PDZ domain and mediates neurexin clustering within the axon (125), which appears to be

necessary for neurexin’s synaptogenic function (88, 124, 204, 205). Recently,

mammalian syd-1 was shown to organize synapses through binding with other active

zone proteins, including liprin-2 and munc18-1 (206). However, it is unclear if

mammalian syd-1 binds to neurexin, as the PDZ domain that is critical for neurexin/syd-1

binding in Drosophila is absent in mammals.

21

Figure 1.3: Recruitment of presynaptic cargo downstream of neurexin-dependent

synaptogenic adhesion. Neurexin interacts with a variety of presynaptic proteins;

however, the mechanisms through which STVs and PTVs are incorporated into the

22

synapse remain unclear. This figure illustrates potential mechanisms that might be

important for this process. (1) Bassoon directly interacts with dynein light chains 1 and 2,

as well as myosin V. Blocking the bassoon/dynein interaction leads to aberrant bassoon

and piccolo synaptic localization. (2) The PDZ-domains of Mint1, Veli/MALS, or CASK

might interact with other synaptic proteins in STVs or PTVs in order to trap them at a

developing synapse. (3) F-actin nucleation mediated by Protein 4.1 might enable the

removal of STVs and/or PTVs from their MT highways via myosin motor binding. (4)

Liprin- can serve as a linker molecule between STVs and the KIF1 motor. Delivery of

liprin- to the synapse might serve to cluster and/or recruit other synaptic components.

Alternatively, modulation of liprin- to reduce its binding to KIF1 might play a role in

cargo delivery. (5) CASK can directly phosphorylate substrates at presynaptic terminals.

Therefore, it might phosphorylate a motor-binding protein integral to STVs or PTVs, a

motor-cargo adaptor or the motor itself, which has been shown to reduce binding between

kinesins and STVs. (6) Although neurexin interacts with a plethora of cytosolic targets,

its cytoplasmic domain is completely dispensable for many of its synaptogenic properties.

This suggests that cis interactions between the extracellular domain of neurexin and a

yet-to-be-determined trans-membrane effector molecule(s) can induce recruitment of

presynaptic proteins. The number and type of downstream recruitment mechanisms

mediated by this potential interaction is unknown.

23

Many additional aspects of trans-synaptic signaling remain unclear. For example,

trans-synaptic signaling does not affect all STVs and PTVs equally: many vesicles can

pass directly through regions of trans-synaptic adhesion without becoming recruited into

a synapse (88, 96). It was also previously not known whether STVs and PTVs were

actively transported to sites of synaptogenic signaling or passively trapped at these sites.

Furthermore it is currently unknown how trans-synaptic adhesion causes STVs and PTVs

to dissociate from their MT motors and become incorporated into the synapse.

Identifying these mechanisms and determining how they are coordinated during

synaptogenesis will be critical to understanding how presynaptic terminals form. To that

extent, experiments described here in Chapters 3 and 4 attempt to elucidate some of these

mechanisms

In this thesis, a number of major issues concerning presynaptic development will

be addressed. In Chapter 2, the trafficking of two transport vesicles carrying distinct sets

of presynaptic proteins was found to be coordinated prior to synapse formation,

uncovering a potential mechanism for rapid recruitment of components into a developing

synapse. The experiments in Chapter 3 determine the dynamics through which

presynaptic proteins arrive at developing synapses and describe potential mechanisms

through which these components become recruited. Finally, in Chapter 4, in vivo imaging

via two-photon microscopy through cranial windows is utilized to determine the

dynamics of presynaptic terminals in the developing mouse brain.

24

Chapter 2: Coordinated trafficking of synaptic vesicle and active zone proteins

prior to synapse formation

Luke A.D. Bury1

and Shasta L. Sabo1,2

Departments of 1Pharmacology and

2Neuroscience, Case Western Reserve University

School of Medicine, Cleveland, Ohio.

*Parts of this chapter were published in the journal Neural Development, Volume 6,

Issue 24, May 10, 2011.

25

Abstract

The proteins required for synaptic transmission are rapidly assembled at nascent

synapses, but the mechanisms through which these proteins are delivered to developing

presynaptic terminals are not understood. Prior to synapse formation, active zone

proteins and synaptic vesicle proteins are transported along axons in distinct organelles

referred to as piccolo-bassoon transport vesicles (PTVs) and synaptic vesicle protein

transport vesicles (STVs), respectively. Although both PTVs and STVs are recruited to

the same site in the axon, often within minutes of axo-dendritic contact, it is not known

whether or how PTV and STV trafficking is coordinated before synapse formation. Here,

using time-lapse confocal imaging of the dynamics of PTVs and STVs in the same axon,

we show that vesicle trafficking is coordinated through at least two mechanisms. First, a

significant proportion of STVs and PTVs are transported together before forming a stable

terminal. Second, individual PTVs and STVs share pause sites within the axon.

Importantly, for both STVs and PTVs, encountering the other type of vesicle increases

their propensity to pause. To determine if PTV-STV interactions are important for

pausing, PTV density was reduced in axons by expression of a dominant negative

construct corresponding to the syntaxin binding domain of syntabulin, which links PTVs

with their KIF5B motor. This reduction in PTVs had a minimal effect on STV pausing

and movement, suggesting that an interaction between STVs and PTVs is not responsible

for enhancing STV pausing. Our results indicate that trafficking of STVs and PTVs is

coordinated even prior to synapse development. This novel coordination of transport and

pausing might provide mechanisms through which all of the components of a presynaptic

terminal can be rapidly accumulated at sites of synapse formation.

26

Introduction

In the cortex, the bulk of synapse formation occurs at high rates over a period of

several weeks during early postnatal development. Formation of individual synapses is

triggered when axons and dendrites contact one another (45, 46). For these axo-dendritic

contacts to stabilize and form a synapse, synaptic vesicle and active zone proteins must

be recruited to the site of contact very rapidly, within minutes to hours (51, 53, 115, 117,

118, 207-209). This rapid assembly is remarkable, considering that each presynaptic

protein needs to be transported from the soma to individual sites of synapse formation

along the axon.

Recent work using live imaging of developing neurons has revealed mechanisms

that might facilitate rapid synapse assembly. For example, rather than transport each

molecule individually, neurons package multiple presynaptic components into vesicles in

the cell body and transport the entire group together. Two major groups of proteins are

transported in this fashion: active zone proteins are transported in piccolo-bassoon

transport vesicles (PTVs), while synaptic vesicle-associated proteins are transported in

synaptic vesicle protein transport vesicles (STVs) (47-51, 53, 59, 60, 210). PTVs and

STVs can be distinguished both morphologically and biochemically. PTVs have been

described as dense-core vesicles or aggregates of vesicles and proteins that range in size

from approximately 80nm in diameter for dense core vesicles to 130nm by 220nm in area

for aggregates (49, 59). Proteins transported by PTVs include piccolo, bassoon, N-

cadherin and syntaxin (49, 59, 60, 73, 74). STVs are heterogeneous in size and shape and

are comprised of both tubulovesicular and clear core vesicles (47, 48, 53). Proteins

carried by STVs include synaptophysin, synapsin Ia, synaptotagmin and

27

synaptobrevin/vesicle-associated membrane protein 2 (VAMP2) (47-49, 51, 53). Both

types of vesicles are packaged via the trans-Golgi network before being transported

through the axon (48, 49, 61).

Rapid assembly of presynaptic terminals can also be facilitated by having the

source of synaptic proteins nearby and readily available. Indeed, many STVs and PTVs

are mobile within axons well before synapse formation. STVs and PTVs travel in a

saltatory fashion in both the anterograde and retrograde directions along the entire length

of the axon [4, 8, 11, 13-15, 17, 19]. This provides a readily-available pool of synaptic

vesicle and active zone proteins wherever and whenever axo-dendritic contact occurs.

Since both STVs and PTVs must be delivered to the same site during synapse

assembly, we hypothesized that trafficking of STVs and PTVs might be coordinated even

prior to synapse assembly, providing an additional mechanism to facilitate rapid

accumulation of the full complement of proteins required for synaptic transmission. Such

coordination could arise through co-transport of STVs and PTVs, through pausing of

STVs and PTVs at the same sites, or both. Coordinated pausing presents a particularly

intriguing possibility since sites along the axon where STVs pause their transport are

preferential sites of synapse formation (51). It was previously suggested that cues that

control pausing at these sites could promote rapid synapse assembly by increasing the

probability that STVs are at or near any given site when axo-dendritic contact occurs. If

this model is correct, then synapse assembly would be most efficient if PTVs are also

attracted to these same sites. However, it is not yet known whether PTVs also pause at

these sites and, if so, whether they do so simultaneous with STVs. A recent report

demonstrated that PTVs and STVs (particularly those resembling small, clear vesicles)

28

can be observed tethered together in electron micrographs of developing axons (49).

These aggregates could correspond to PTVs and STVs that are either being co-

transported or pausing at the same site at the same time.

Here, we tested whether transport and pausing of PTVs and STVs is coordinated

prior to synapse assembly using time-lapse confocal imaging of green and red fluorescent

protein-tagged synaptic vesicle and active zone proteins within the same axon. We found

that a significant portion of STVs and PTVs move together and share pause sites.

Interestingly, both STVs and PTVs preferentially paused at these sites when another

vesicle was present. These observations raised the question of whether a direct interaction

between STVs and PTVs coordinates their transport and pausing. Reducing PTVs in the

axon minimally affected the movement and pausing of STVs, arguing against this

mechanism and suggesting that other unidentified signals are responsible for STV and

PTV coordination. These findings represent novel mechanisms that can facilitate the

rapid recruitment of presynaptic proteins to the same site within the axon and, therefore,

promote synapse development.

29

Results

STVs and PTVs can move together

STVs and PTVs move in a similar saltatory fashion within the axon in both the

anterograde and retrograde directions (47, 48, 51, 53, 58, 60, 73). However, because

STVs and PTVs have almost always been imaged separately, it is not yet known if and

how STV and PTV trafficking interrelate. To determine this, we used time-lapse imaging

of neurons co-transfected with fluorescent STV and PTV markers. In this assay, 5-6 DIV

rat cortical neurons were co-transfected with GFP-bassoon to label PTVs and either

synaptophysin-mcherry or synaptophysin-mRFP to label STVs. Previous studies have

indicated that fluorescent proteins attached to bassoon and synaptophysin effectively

label PTVs and STVs, respectively (48, 50, 53, 60, 61, 115, 210, 211). 1-2 days after

transfection, time-lapse images of the fluorescently labeled vesicles were obtained from

individual axons. Images were taken every 10s for 7.5 minutes.

From the kymographs in Figure 2.1A, it is clear that PTVs and STVs moved

together in both anterograde and retrograde directions. To quantify these correlated

movements, the positions of individual STV and PTV puncta were tracked independently

while blind to the other channel, and then their movements were compared. To

determine whether these correlated movements could be accounted for by chance, the

experimental data were ultimately compared to model axons where the initial positions of

all of the puncta were randomized but the density of puncta and properties of their

movements were derived from the experimental data (Figure 2.1B).

30

During the imaging period, 18.3% of PTVs moved with STVs (n = 224 vesicles),

while 37.1% of STVs moved with PTVs (n = 97 vesicles; Figure 2.1C). STVs and PTVs

that moved together also spent the majority of their time together (Figure 2.1D). The

prevalence of correlated STV and PTV movements was over 10-fold higher for the

experimental data than predicted by the same analysis of model axons (Figure 2.1C;

model: n = 22400 PTVs and 9700 STVs). In addition, the percentage of time STVs and

PTVs spent moving together was over 2.5-fold higher expected by chance (Figure 2.1D).

These data indicate that a population of STVs and PTVs travel together within the axon,

which could allow both sets of presynaptic proteins to be distributed together to sites of

synapse formation.

31

Figure 2.1: STVs and PTVs move together. (A) Kymographs of an axon segment

showing the movements of PTVs (GFP-bassoon; green) and STVs (synaptophysin-

mcherry; magenta) in the same axon. Yellow lines underline the moving STVs and PTVs.

Bottom panel, overlay of STV and PTV fluorescence. On the ordinate axis, one pixel

corresponds to 10s. On the abscissa, the scale bar corresponds to 10µm. (B) Simulations

of model axons were performed by randomizing initial positions of vesicles while

maintaining movement and pausing characteristics of the original imaged vesicles.

Diagrams illustrate kymographs of 3 simulations of model axons. Movements of two

32

vesicles are shown in each model axon. (C) Plot showing the percentage of PTVs that

move with STVs (green) and STVs that move with PTVs (magenta). PTV and STV

simulations (light green and magenta, respectively) correspond to the predicted values

from the simulations. (D) Quantification of the percentage of time that PTVs and STVs

spent together. Both the percentage of vesicles that moved together and the time spent

together were substantially larger in the data than is predicted by chance (simulations).

33

STVs and PTVs display similar pausing characteristics

Although the movements of both STVs and PTVs have been extensively

described individually, the pauses between movements have been less thoroughly

analyzed and have not been compared. Comparing STV and PTV pausing behavior is

important since STVs preferentially pause at sites of eventual synapse formation (51, 53).

When PTV behavior was examined, it was clear that PTV pausing exhibited many of the

properties previously described for STVs (Figure 2.2A-B and (51)). For example, the

same PTV could be observed repeatedly pausing at a given site (Figure 2.2B, top panel).

In addition, multiple PTVs paused at the same site, both sequentially and simultaneously

(Figure 2.2B, middle and bottom panels, respectively). When STV movements were

quantified, the mean STV pause frequency was 0.0081 +/- 0.0002 pauses/second (n = 262

vesicles, Figure 2.2C). The average STV pause duration was 107.4 +/- 3.4 seconds (n =

954 pauses, Figure 2.2D). Both measurements concur with previous observations (51).

As shown in Figure 2.2C-D, PTV pause frequency and duration in 7-8 DIV neurons were

nearly identical to the STV data, with PTVs pausing at a frequency of 0.0082 +/- 0.0003

pauses/second and average duration of 108.3 +/- 3.5 seconds (n = 253 vesicles, n = 933

pauses). The similarities between the pausing behavior of STVs and PTVs raise the

question of whether PTVs might pause with STVs at sites of eventual synapse formation.

Like STVs (51), regulation of PTV pausing could be an important target of signals that

control synapse assembly.

34

Figure 2.2: PTV pausing is qualitatively and quantitatively similar to STV pausing.

(A) Time-lapse images of PTVs labeled with GFP-bassoon. Individual PTVs are tracked

with red, yellow or blue arrows. Most PTVs paused, and pauses were of varied durations.

Frames were collected at the indicated times. Scale bars, 5 µm. (B) Kymographs

35

demonstrating the movements of PTVs in 3 axons. Individual PTVs paused repeatedly at

the same site (top panel), and multiple PTVs paused at the same site (middle and bottom

panels). Pausing of multiple PTVs at a given site occurred both sequentially (middle

panel) and simultaneously (bottom panel). Individual vesicles are highlighted in different

colors for visualization. Pause sites are outlined in orange. On the ordinate axis, one pixel

corresponds to 10s. (C-D) STVs (magenta) and PTVs (green) paused at similar

frequencies (C) and for similar mean durations (D). Data represent the mean + S.E.

36

STVs and PTVs pause simultaneously at the same sites in the axon

For a synapse to develop, both active zone and synaptic vesicle proteins need to

be recruited to the same site. Sharing a pause site provides a potential mechanism for

recruiting both STVs and PTVs to the same spot in the axon. Since places where STVs

pause are preferred sites of synapse formation, it seemed likely that PTVs also pause at

these sites. To test this hypothesis, STVs and PTVs were imaged and analyzed as

described above. From the time-lapse panels and kymographs shown in Figure 2.3A-B, it

is clear that both types of vesicles paused at sites where the other type of vesicle also

paused.

To determine whether pausing of STVs and PTVs at the same sites was

significantly greater than would be predicted by chance, pausing at the same sites was

quantified in each axon then compared to simulations in model axons (Figure 2.3C). For

STVs, 47.9% of pauses were at sites where a PTV also paused (n = 305 pauses, Figure

2.3D), compared to 24.5% using the same analysis of model axons (n = 33967 pauses,

Figure 2.3D). This indicates that STVs preferentially pause at PTV pause sites.

Likewise, 24.5% of all PTV pauses were at sites where an STV also paused (n = 620

pauses, Figure 2.3D), compared to 14.0% using the same analysis of model axons (n =

69263 pauses, Figure 2.3D), indicating that PTVs preferentially pause at STV pause sites.

Some STVs do not encounter PTV sites during our imaging time window and vice versa.

To account for this, we also quantified pausing at the same site with the analysis limited

to STVs and PTVs that had the opportunity to pause at PTV and STV sites, respectively

(Figure 2.3E). 57.8% of STVs that encountered a PTV site paused (n = 135), and 84.4%

of PTVs that encountered STV sites paused at those sites (n = 90). PTVs are slightly

37

more likely to pause at STV sites; however, this difference appears to be an inherent

property of the vesicles and is accounted for in model axons (see Figure 2.4A,C). It has

been shown previously that STV pause sites are preferred sites of synapse formation, and

PTVs pause at these same sites; therefore, PTVs pause at sites of synapse formation.

38

Figure 2.3: STVs and PTVs pause at the same sites. (A) Time-lapse images of a

segment of axon expressing both synaptophysin-mcherry (magenta) and GFP-bassoon

39

(green). In each panel, fluorescence signals from the two channels are overlaid, and

colocalization is indicated by white. Orange box, site where a PTV is paused. The arrow

tracks a STV that then pauses at that site. Scale bars, 5 µm. (B) Kymographs showing

two examples of STVs and PTVs that pause at the same sites. On the ordinate axis, one

pixel corresponds to 10s. Bottom panel, overlay of STV and PTV fluorescence. Orange

boxes outline 3 shared pause sites. (C) Simulations of model axons were performed by

randomizing initial positions of vesicles while maintaining movement and pausing

characteristics observed for the original experimental vesicles. Diagrams show the

superimposed locations of all pause sites for individual STVs and PTVs in model axons.

Three simulations of the same experimental data are shown. STV pause sites are

indicated with a dot while PTV pause sites are indicated with a “+” sign. Each color

represents an individual vesicle. Cyan and magenta boxes outline the pause sites of the

same PTV and STV, respectively, in each model axon. For each axon imaged, 100 such

simulations were performed, allowing estimation of the degree of co-transport and co-

pausing expected from chance alone. (D) Plot illustrating the percentage of PTV (green)

and STV (magenta) pause sites that are shared with STVs and PTVs, respectively. The

fraction of shared sites is much higher than predicted by chance (via simulations, light

green and light magenta). (E) A large majority of PTVs that encountered STV pause

sites paused at those sites. Similarly, most STVs that encountered PTV pause sites then

paused at those sites. Error bars display the 95% confidence interval.

40

Since both PTVs and STVs pause in the same places, the next question addressed

was whether STVs and PTVs stop at these sites at the same time. Synapse assembly is a

rapid process, and concurrent attraction of both STVs and PTVs could provide a

mechanism for efficient assembly of presynaptic terminals. To determine whether PTVs

and STVs are simultaneously attracted to sites of synapse formation, we quantified

whether STVs preferentially pause at PTV pause sites when a PTV is present. As shown

in Figure 2.4, STVs were more likely to pause at a site that contained a PTV than a site

that did not. When STVs encountered pause sites with PTVs present, 76.4 +/- 11.2% of

STVs paused with the PTV (n = 55; Figure 2.4A). In contrast, only 45.0 +/- 10.9% of

STVs paused at PTV pause sites when a PTV was not there (n = 80). This 1.7-fold

difference in the likelihood of pausing with a PTV could not be accounted for by chance

because there was no interaction between STVs and PTVs when the same analysis was

performed on data obtained from simulations with model axons (Figure 2.4A). In the

model, 52.4 +/- 1.9% of STVs paused at PTV pause sites with a PTV present (n = 2334),

and 51.6 +/- 1.3% of STVs paused at PTV pause sites without PTVs present (n = 5767).

The pause duration of an STV stopped at a PTV pause site was also measured. In

contrast to the likelihood of pausing, the pause duration of STVs is not affected by the

presence or absence of PTVs (Figure 2.4B). The average pause duration for STVs

paused at sites with PTVs present was 101.0 +/- 13.4 seconds (n = 42 pauses), while the

average pause duration for an STV paused at a site without a PTV present was 94.2 +/-

20.1 seconds (n = 36).

It is not yet known whether STVs and PTVs must be recruited to nascent

presynaptic terminals in a defined order, so we also tested whether PTVs are more likely

41

to pause at sites that contain STVs. As shown in Figure 2.4C, PTVs paused at sites that

contained STVs at a significantly higher rate (93.6 +/- 7.0%, n = 47) than at sites where

STVs had previously paused but were no longer present (74.4 +/- 13.0%, n = 43). In

contrast, this effect of STVs on PTV pausing behavior was not seen with analysis of the

randomized model: in the simulations, 76.0 +/- 2.0% of PTVs paused at STV sites with

an STV present (n = 1591), and 74.7 +/- 1.1% of PTVs paused at STV sites without

STVs present (n = 4906). There was no difference in the probability that either vesicle

arrived first at a shared pause site, based on 95% CI (data not shown). Similar to STVs

pausing at PTV sites, the mean pause duration of PTVs pausing at STV pause sites is not

affected by the presence of STVs (Figure 2.4D). The average pause duration for PTVs

paused at sites with STVs present was 100.5 +/- 12.4 seconds (n = 44 pauses), while the

average pause duration for PTVs paused at sites without STVs present was 80.9 +/- 15.3

seconds (n = 32 pauses). These data demonstrate that PTVs prefer to pause at sites that

contain STVs and vice versa, suggesting that STVs and PTVs are attracted to the same

sites at the same time.

42

Figure 2.4: STVs and PTVs preferentially pause at the same sites at the same time.

(A) STVs (magenta) are significantly more likely to stop at a PTV pause site when PTVs

are at the site. This preference cannot be accounted for by chance since no dependence on

the presence of PTVs was seen in simulations (light magenta). The data presented

correspond to the mean + 95% confidence intervals. *, 95% confidence intervals do not

43

overlap. (B) STVs pause for similar lengths of time regardless of whether a PTV is

present at the pause site. Data are presented as mean + S.E. (C) PTVs (green) are more

likely to pause at a site when an STV is present. The same dependence was not observed

in simulations (light green). (D) PTVs pause for similar lengths of time, regardless of

whether STVs are present at the pause sites. Data are the mean + S.E.

44

Recruitment of a PTV enhances accumulation of additional PTVs at sites of synapse

formation

Assembly of presynaptic terminals involves recruitment of multiple PTVs (60).

Therefore, we wondered whether PTVs might also increase attraction of other PTVs to

sites of synapse formation. To test this, we quantified the percentage of PTVs that

paused at sites where another PTV was either present or had previously paused. When a

PTV was already anchored at a particular pause site, there was high likelihood of an

additional PTV pausing at that site: 83.3 +/- 13.3% of PTVs paused when they

encountered sites with other PTVs (n = 30; Figure 2.5A). However, when a PTV was not

present at the pause site, the likelihood that a PTV would pause was significantly lower

(42.5 +/- 15.3%, n = 40). This nearly 2-fold increase in attraction of PTVs to sites

containing other PTVs could not be explained by chance since the increase was not

observed in our simulations. In model axons, PTVs paused at 67.1 +/- 1.6% of sites with

other PTVs (n = 1267) and 68.1 +/- 1.2% of sites without other PTVs (n = 5827; Figure

2.5A). This indicates that the presence of a PTV at a pause site promotes recruitment of

additional PTVs to that site.

Interestingly, the presence of an STV at a particular pause site did not

significantly increase the probability of an additional STV pausing at that site (chance of

pausing with another STV present = 71.4 +/- 23.7%, n = 14; without another STV present

= 60.5 +/- 15.5%, n = 38; Figure 2.5B). As expected, no dependence of STV pausing on

the presence of other STVs was observed in our randomized model axon (STV present =

56.8 +/- 2.4%, n = 1665; STV absent = 59.8 +/- 1.6%, n = 3443; Figure 2.5B). These data

45

suggest that STVs do not interact with one another in a way that promotes recruitment to

sites of synapse formation.

46

Figure 2.5: Multiple PTVs are attracted to the same sites. (A) PTVs are more likely to

pause when other PTVs are present (green). Spatially randomized simulations are shown

in light green and cannot account for the tendency of PTVs to pause simultaneously with

other PTVs. (B) STVs are not more likely to pause when other STVs are already present

(magenta). Simulations are shown for comparison (light magenta). Error bars represent

95% confidence intervals. *, 95% confidence intervals do not overlap.

47

STV pausing is only mildly dependent on PTVs

The data presented above demonstrate that pausing of STVs at sites of synapse

formation is enhanced at sites that contain PTVs and vice versa. This raises the question

of whether a direct physical interaction between STVs and PTVs is important for

recruitment of synaptic proteins to sites of synapse formation. If so, then STV pausing

should be altered if PTVs are decreased or eliminated from the axon. Conveniently, a

method for disrupting PTV transport – and, therefore, decreasing PTV density in the axon

-- has recently been described (73, 74). This approach utilizes a dominant negative

construct that disrupts the connection between the PTV-associated protein syntaxin and

syntabulin, which links PTVs to the kinesin motor KIF5B. This construct mimics the

syntaxin binding domain (SBD) of syntabulin and interferes with the syntaxin-syntabulin

interaction. Expression of syntabulin-SBD has been shown to specifically prevent the

majority of PTVs from being transported out of the cell body and into the axon without

directly affecting STV transport or other KIF5 cargo (Figure 2.6A; (73, 74)). Imaging of

our cortical cultures confirmed that the density of endogenous piccolo puncta is

dramatically decreased in axons of neurons expressing syntabulin-SBD fused to GFP

when compared to neurons expressing GFP alone (Figure 2.6B).

Consistent with the published data, transfection of neurons with the syntabulin-

SBD construct did not interfere with STV transport in axons (Figure 2.6C): STVs still

moved, and the mean distance moved over the course of imaging was unchanged (No

SBD = 8.90 +/- 0.93 µm, n = 163 vesicles; SBD = 9.30 +/- 0.82 µm, n = 153 vesicles, p =

0.68). When STV pausing was quantified, expression of the syntabulin-SBD construct

yielded no change in the pause frequency (No SBD = 0.0085 +/- 0.0003 pauses/second, n

48

= 163 vesicles; SBD = 0.0080 +/- 0.0004 pauses/second, n = 153 vesicles; p = 0.62;

Figure 2.6D) but did cause a slight decrease in the average duration of STV pauses (No

SBD = 109.9 +/- 4.4 s, n = 586 pauses; SBD = 102.4 +/- 4.3 s, n = 582 pauses; p = 0.03;

Figure 2.6E). In addition, the instantaneous velocities of STVs were slightly reduced in

SBD-transfected neurons (No SBD = 0.208 +/- 0.004 µm/s, n = 729 movements; SBD =

0.193 +/- 0.004 µm/s, n = 774 movements; p = 0.003; Figure 2.6F). Although there was a

decrease in STV pause duration when PTV transport was disrupted, the magnitude of the

effect and the lack of a change in the probability of pausing argue against PTVs

themselves controlling STV pausing. These data suggest that a direct interaction between

STVs and PTVs is not responsible for the increased attraction of STVs to sites that

contain PTVs.

49

50

Figure 2.6: A direct interaction between STVs and PTVs cannot account for the

attraction of STVs to pause sites that contain PTVs. (A) PTV transport was inhibited

using dominant-negative syntabulin (syntaxin binding domain, SBD) fused to GFP. (B)

Expression of syntabulin-SBD-GFP decreases the density of PTV puncta in axons when

compared to axons expressing GFP alone. PTVs were identified by immunofluorescent

labeling for endogenous piccolo. (C) Images and kymograph (bottom) showing that STVs

(magenta) move and pause in SBD-expressing axons (green). (D) The frequency of STV

pausing was unchanged in the presence of SBD-GFP. (E) STVs pause durations are

shorter when PTV localization in the axon is disrupted. p-values are from Wilcoxon rank-

sum test. The change in pausing upon expression of SBD-GFP is not sufficient to account

for the coordinated pausing of STVs and PTVs. (F) The instantaneous velocities of STVs

are increased in SBD-expressing axons. Black arrow, mean instantaneous velocity.

51

Materials and Methods

All studies were conducted with an approved protocol from the Case Western

Reserve University Institutional Animal Care and Use Committee, in compliance with the

National Institutes of Health guidelines for the care and use of experimental animals.

Neuronal cultures and transfection

Primary neuronal cultures were prepared from postnatal rat visual cortex

essentially as described previously [4, 5, 15], except neurons were maintained in

Neurobasal-A medium supplemented with glutamax and B27 (Invitrogen). Neurons were

transfected with Lipofectamine 2000 (Invitrogen, Carlsbad, CA) 24-48 h before live

imaging. Excluding GFP-bassoon transfection, 1µg of DNA construct was combined

with 1µg of Lipofectamine 2000 in 50ul of Optimem (Invitrogen) for each 18 mm

coverslip. Transfection of GFP-bassoon was conducted in the same manner except 2µg

of DNA were used to account for its large size. With double transfection, localization of

each protein appeared similar when expressed alone. Also, the distribution and

movement of STVs labeled with synaptophysin-mcherry, synaptophysin-mRFP, and

synaptophysin-GFP all appeared similar to one another. GFP -bassoon (GFP-Bsn 95-

3938), synaptophysin-mcherry, synaptophysin-mRFP, synaptophysin-GFP and syntaxin-

SBD-GFP were generous gifts of Drs. Thomas Dresbach (University of Heidelberg,

Heidelberg), Matthijs Verhage (Vrije Universiteit, Amsterdam), Jurgen Kingauf

(University of Muenster, Muenster), Jane Sullivan (University of Washington, Seattle)

and Zu-Hang Sheng (National Institute of Neurological Disorders and Stroke, Bethesda).

52

Each of these constructs has been shown previously to be functional and properly

localized (50, 61, 73, 211-213).

Live imaging

Neurons were imaged at 7-8 DIV with a C1 Plus confocal system with a Nikon

Eclipse Ti-E microscope using a 40x Nikon Plan Apo 0.95 numerical aperture objective.

Lasers were 488 nm argon and 543 nm helium-neon. Detection filters were 515/30 nm

bandpass for GFP and 590/50 nm bandpass for mcherry/mRFP. Images were collected

every 10s, with scan times no greater than 3.3s. This imaging interval was selected as a

compromise between having a high temporal resolution and minimizing the time the

neurons were exposed to the laser to avoid any toxicity and photobleaching. A total of 45

images were collected for each time-lapse series. Imaging was conducted with constant

perfusion with artificial CSF (ACSF) (120 mM NaCl, 3 mM KCl, 2 mM CaCl2, 2 mM

MgCl2, 30 mM D-glucose, 20 mM HEPES, and 0.2% sorbitol, pH 7.3). ACSF perfusion

was performed at 25oC since STV transport and pausing are not significantly different at

ambient and physiological temperatures (51). Axons were identified using morphological

criteria, as described previously (50, 51). For dual-color imaging, channels were

collected sequentially to eliminate bleed-through and neurons were imaged in which

expression levels of both fusion proteins appeared comparable.

Immunofluorescence

Neurons were fixed for 15 minutes in 4% paraformaldahyde in PBS containing

4% sucrose, permeabilized for 5 minutes with 0.2% Triton X-100, and blocked with 10%

horse serum. The primary antibody was piccolo (Synaptic Systems), and the secondary

53

antibody was Alexa 633-conjugated goat anti-rabbit (Invitrogen). Coverslips were

mounted in Fluoromount (Fisher Scientific, Pittsburgh, PA) containing DABCO (1,4-

diazabicyclo[2.2.2]octane) (Sigma, St. Louis, MO).

Analysis and statistics

STV and PTV movements were tracked using ImageJ (NIH, Bethesda, MD). To

restrict the quantification of time-lapse movies to healthy neurons, axons were included

in the analysis only if a least 1 vesicle moved within the field of view. Regions of axon

with an intermediate STV or PTV density were imaged to allow us to track each vesicle

reliably. STV and PTV movements were tracked independently while blind to the other

channel.

For quantification of movement and pausing, positional data were imported into

Matlab and analyzed with custom-written programs (Mathworks, Natick, MA). For

pause analysis, a pause was defined as a period greater than or equal to 10s, during which

the velocity of the vesicle went to 0+/- 0.1 µm/s. The cutoff zero velocity (0.1µm/s) was

chosen based on the average diameter of vesicles, which was approximately 1µm.

Vesicles that never moved were not included in the analysis. Therefore, vesicles which

pause for very long durations might be underrepresented. Also, only vesicles that could

be tracked for the entire imaging duration were included in the analysis. Vesicles that

move at high velocities are more likely to leave the imaging area before the end of the

movie. Therefore, these vesicles might also be underrepresented in the analysis.

Given the pixel density, scan speed and average vesicle size, a vesicle was

typically imaged in approximately 10-50 ms, permitting reliable tracking of vesicles

54

based on their apparent size, shape, and intensity, with relatively low influence of vesicle

movements on these parameters. The fastest STV movements, at a maximum

approaching 1 µm/s (comparable with the fastest velocities that have been reported for

STVs (51, 53, 210)), resulted in movements of 10 µm during the imaging interval and

were easily measured. Maximal velocities of PTVs are lower than maximal STV

velocities (60, 73) and were also easily recorded.

Data are presented cumulatively with 95% confidence intervals for binomially

distributed data or as the mean +/- standard error of the mean (S.E.) where appropriate.

Confidence intervals were calculated based on a normal approximation, and data sets

were considered significantly different if their 95% confidence intervals were non-

overlapping. For data presented as means, significance was evaluated using the Wilcoxon

rank sum test.

Randomized model

In some cases, experimental data were compared to randomized simulations of

vesicle movement. For each axon that was imaged, a model axon was generated with the

same number of vesicles and axonal length as each experimental axon. In these models,

the initial position of each vesicle was randomized. However, the number, timing and

direction of movements, as well as the instantaneous velocity, and pause duration of each

vesicle remained the same. The newly generated time-lapse series were then analyzed in

the same manner as the experimental data. Each simulation was performed 100 times per

axon to enhance statistical analysis.

55

Acknowledgements

We thank Michael Sceniak for advice and help with Matlab programming, for help with

design of the simulations and for helpful discussions about the project. We also thank

Corbett Berry and Louie Zhou for technical assistance. We are grateful to Drs. Thomas

Dresbach (University of Heidelberg, Heidelberg), Matthijs Verhage (Vrije Universiteit,

Amsterdam), Jurgen Kingauf (University of Muenster, Muenster), Jane Sullivan

(University of Washington, Seattle) and Zu-Hang Sheng (National Institute of

Neurological Disorders and Stroke, Bethesda) for GFP-bassoon (GFP-Bsn 95-3938),

synaptophysin-mcherry, synaptophysin-mRFP, synaptophysin-GFP and syntaxin-SBD-

GFP constructs, respectively. This work was funded by a Research Starter grant from the

PhRMA Foundation (SS), start-up funds from Case Western Reserve University (SS),

and the NIH Predoctoral Training Program in Molecular Therapeutics at Case Western

Reserve University (LB).

56

Chapter 3: Dynamic mechanisms of neuroligin-dependent presynaptic terminal

assembly in living cortical neurons

Luke A.D. Bury1

and Shasta L. Sabo1,2

Departments of 1Pharmacology and

2Neuroscience, Case Western Reserve University

School of Medicine, Cleveland, Ohio.

*Parts of this chapter were published in the journal Neural Development, Volume 9,

Issue 13, May 29, 2014.

57

Abstract

Synapse formation occurs when synaptogenic signals trigger coordinated

development of pre and postsynaptic structures. One of the best-characterized

synaptogenic signals is trans-synaptic adhesion. However, it remains unclear how

synaptic proteins are recruited to sites of adhesion. In particular, it is unknown whether

synaptogenic signals attract synaptic vesicle (SV) and active zone (AZ) proteins to

nascent synapses or instead predominantly function to create sites that are capable of

forming synapses. It is also unclear how labile synaptic proteins are at developing

synapses after their initial recruitment. To address these issues, we used long-term, live

confocal imaging of presynaptic terminal formation in cultured cortical neurons after

contact with the synaptogenic postsynaptic adhesion proteins neuroligin-1 or SynCAM-1.

Surprisingly, we find that trans-synaptic adhesion does not attract SV or AZ proteins nor

alter their transport. In addition, although neurexin (the presynaptic partner of neuroligin)

typically accumulates over the entire region of contact between axons and neuroligin-1-

expressing cells, SV proteins selectively assemble at spots of enhanced neurexin

clustering. The arrival and maintenance of SV proteins at these sites is highly variable

over the course of minutes to hours, and this variability correlates with neurexin levels at

individual synapses. Together, our data support a model of synaptogenesis where

presynaptic proteins are trapped at specific axonal sites, where they are stabilized by

trans-synaptic adhesion signaling.

58

Introduction

Upon axon-dendrite contact, the extracellular domains of axonal and dendritic

adhesion proteins interact, leading to recruitment of synaptic proteins and subsequent

synapse formation (5, 136, 138). A variety of synaptogenic adhesion partners have been

discovered, including neurexin/neuroligin, SynCAM/SynCAM, and neurexin/LRRTM2,

among others (117, 128, 129, 143, 163-166, 176, 214-216). Of these, the most studied

trans-synaptic pair is postsynaptic neuroligin and its interaction with presynaptic

neurexin.

Although a number of studies have focused on understanding the functions and

mechanisms of trans-synaptic adhesion molecules in recent years, it remains unclear how

synaptogenic adhesion results in presynaptic protein recruitment and synaptogenesis

(124, 217). Importantly, it is not known whether trans-synaptic adhesion actively attracts

synaptic proteins to sites of signaling or acts primarily to stabilize proteins at nascent

terminals. In addition, it is unknown whether synaptic protein recruitment proceeds

continuously until the site is saturated, or if SV and AZ protein levels are modulated even

at the earliest stages of development, as they are at mature synapses (218-225). Finally, it

is unclear if trans-synaptic adhesion regulates synaptic protein levels at individual

synapses or primarily creates sites in the axon that are capable of synapse formation

while other cellular processes regulate levels of recruitment.

To answer these questions, we induced synaptogenesis via contact with

neuroligin-1 and SynCAM-1, then employed short and long-term time-lapse confocal

imaging of presynaptic protein recruitment to nascent sites of trans-synaptic signaling.

59

This paradigm allowed us to record recruitment from the initiation of trans-synaptic

adhesion onward. We found that SV protein recruitment to individual sites of trans-

synaptic adhesion fluctuated, as did protein recruitment at axon-dendrite synapses. These

changes in synaptic protein levels strongly correlated with the amount of neurexin-1β at

individual synapses. However, trans-synaptic adhesion did not attract SV or AZ proteins

to contacts or significantly alter synaptic protein transport. Unexpectedly, neurexin alone

was not the primary signal for recruitment since synaptic proteins were specifically

recruited to sites of enhanced neurexin-1β clustering, even though neurexin was present

throughout the contact region. Overall, our data support a model of synaptogenesis in

which presynaptic proteins are trapped at, but not actively attracted to, specific axonal

sites, where they are stabilized by trans-synaptic adhesion signaling.

60

Results

Synaptic protein recruitment fluctuates frequently and rapidly during assembly of

individual presynaptic terminals

Contact between an axon and dendrite is one of the first steps of synapse

formation; however, it is difficult to predict when and where axo-dendritic contact will

occur and whether synapses will form at these sites. To overcome this challenge, we

developed a strategy for long-term imaging of synaptic protein recruitment to nascent

sites of presynaptic terminal formation, initiated by contact with neuroligin-1-expressing

HEK293 cells as a proxy for post-synaptic dendrites. Specifically, 7-10DIV cortical

neurons were sparsely transfected with synaptophysin-GFP, to label synaptic vesicle

protein transport vesicles (STVs), which deliver SV proteins to developing presynaptic

terminals (47-49, 51-53, 58, 68-71, 81, 87, 114, 115, 117, 210, 226, 227).We then imaged

STVs in individual axons that contacted neuroligin-1-expressing HEK293 cells for up to

25h after contact (Figure 3.1A), collecting multiple “short sequences” of rapid imaging

(every 10s for 7.5min) separated by 1-2.5h. The “short sequences” granted the ability to

track rapid STV movements, observe recruitment of individual STVs, and distinguish

between STVs that were stably recruited and those that were en route through the contact

site while preserving neuron health. With this approach, it was possible to resolve a full

time-course of presynaptic protein recruitment at individual sites of trans-synaptic

signaling in individual axons.

To quantify STV recruitment within individual axons, the integrated densities of

all STVs were summed in each axonal region that made contact with a neuroligin-1-

61

expressing HEK293 cell and compared to regions without contact in the same axon.

Since integrated density corresponds to the sum of pixel intensities, STVs with higher

integrated densities were brighter and/or larger, indicating more SV protein. Enrichment

was calculated by determining the difference between fluorescence (integrated density

per length of axon) inside and outside the contact region, divided by the sum of the

fluorescence inside and outside. Overall, in regions that displayed recruitment (see

Methods), synaptophysin-GFP was increasingly enriched at sites of trans-synaptic

adhesion over the first 10h of imaging (Figure 3.1B; n = 12 contacts), similar to previous

observations using immunocytochemistry to examine populations of axons (115). No

enrichment was observed over the same time course when axons were contacted with

HEK293 cells expressing only HcRed, with no neuroligin, indicating that the observed

recruitment was a specific result of trans-synaptic adhesion (Figure 3.1C; n = 5 contacts).

Enrichment at neuroligin contacts was due to increased SV protein recruitment to

contacts without substantial changes in SV protein outside areas of contact (Figure 3.1D).

Within contact areas, SV protein levels started to rise 2h after contact was initiated. This

“rising/recruitment phase” progressed until reaching a plateau 10h after contact initiation

(Figure 3.1D). However, recruitment at individual contacts was quite labile (Figure 3.1E).

Throughout the imaging period, absolute levels of recruited synaptophysin-GFP

increased or decreased from one hour to the next, even for axons that displayed

enrichment for most or all of the 24h imaging period. These substantial fluctuations

occurred during both rising and plateau phases. Importantly, variability in fluorescence

was not caused by focal drift, since we employed Perfect Focus correction to maintain the

focus (see Methods). Therefore, in individual axons, recruitment of proteins does not

62

consistently increase: rather, levels of recruited SV proteins fluctuate while remaining

elevated compared to neighboring axonal areas without trans-synaptic signaling.

63

Figure 3.1: Levels of synaptic vesicle protein enrichment at individual trans-

synaptic adhesion sites are modulated throughout recruitment. (A) Images from live,

long-term, time-lapse confocal imaging of an axon expressing synaptophysin-GFP

(green) and contacting an HEK293 cell (magenta and white outline) that expresses

neuroligin-1 + HcRed. The top panel shows an axon and HEK293 cell at 1h after contact.

In this panel, synaptophysin-GFP at the contact site appears white in the overlay. The

remaining panels show only synaptophysin fluorescence for clarity. Images were

collected 1h, 12.5h and 20h after contact was induced (scale bar = 5m). (B) Time course

of STV enrichment in axonal regions that contact neuroligin-1-expressing cells.

Enrichment corresponds to the difference between the total STV integrated density within

the contact region and outside of the contact, normalized to the total STV integrated

density throughout the axon and expressed as a percentage. Positive values indicate

enrichment inside the contact area. Individual points (cyan) represent mean values from

45 images collected at 10s intervals, beginning at the time indicated on the x-axis.

Dashed line, overall mean at each time point (n = 12 contacts). Overall, enrichment

64

increased gradually over the first 10h then remained elevated. Error bars = S.E.M. (C)

Over the same time course, STVs were not enriched at sites of contact between axons and

HEK293 cells expressing HcRed but no neuroligin. (D) Integrated density of

synaptophysin-GFP inside (solid line) and outside (dashed line) of contacts over time.

Red bar, rising phase of enrichment (2-7.5h after contact). Synaptophysin increased at

contacts while remaining stable outside contacts. (E) Enrichment for two axons. At

individual contacts, enrichment was highly dynamic throughout the imaging period.

Therefore, although STV enrichment increases at trans-synaptic adhesion sites overall,

enrichment levels for individual axons vary throughout the first 24h of development.

65

Fluctuations in recruited SV protein levels could occur through at least two

mechanisms (Figure 3.2A). Once formed, sites of STV recruitment could be stable while

the level of protein at each stable site fluctuates. Alternatively, the sites themselves could

form and disappear. These two mechanisms are not mutually exclusive, and our live

imaging revealed both types of variability. Synaptophysin-GFP levels fluctuated at

individual sites that were stable over several hours (Figure 3.2B). Fluctuations often

occurred very rapidly, within minutes. In addition, sites of stable recruitment often

persisted for multiple hours before completely, and sometimes immediately, disbanding

(Figure 3.2C). Instant formation of stable sites was also observed (Figure 3.2C).

To assess the prevalence of fluctuations in synaptophysin recruitment levels at

individual sites, we determined the number of sites of stable STV accumulation that

displayed fluctuations in synaptophysin-GFP levels over time. Stable accumulation sites

were defined as sites in the axon where synaptophysin-GFP signal was consistently

present over the course of at least one short imaging sequence. At 24.9% of individual

stable accumulation sites (n = 115), there were observable fluctuations of synaptophysin-

GFP within short imaging sequences. For stable accumulation sites that persisted over

multiple short imaging sequences, synaptophysin-GFP levels changed between short

imaging sequences 18.4% of the time (n = 103). It should be noted that these are most

likely conservative estimates of the actual synaptic vesicle protein level fluctuations at

stable sites, as small fluctuations of protein at these sites are difficult to observe.

Additions and losses of entire stable sites of recruitment resulted in net changes in

the density of discrete, stable recruitment sites from one short sequence to the next (i.e.

over 1-2.5 hours) in 76.6% of imaging sequences during the rising phase and 73.8%

66

during the plateau phase. This indicates that sites of recruitment were highly labile

throughout synapse assembly. During the rising phase of recruitment, the additions and

losses resulted in net increases in the density of stable sites of recruitment in 54.7% of

imaging sequences (Figure 3.2D; n = 64). In contrast, during the plateau phase,

appearances and disappearances were more balanced, with net increases in discrete

contact sites only 38.1% of the time (Figure 3.2D).

67

68

Figure 3.2: Modulation of synaptic vesicle protein recruitment occurs through two

distinct mechanisms. (A) Model kymographs illustrating that variability in protein levels

could arise from fluctuations in synaptic protein levels at individual recruitment sites

(left) or through the addition/subtraction of entire recruitment sites (right). (B, C)

Kymographs of STVs (green) in axons that contact neuroligin-1-expressing HEK-293

cells (magenta). (B, box) Synaptophysin-GFP levels at stable contacts are variable over

the course of hours and even minutes (asterisks). (C, top box, top line) Sites of

recruitment that are stable over the course of hours can be eliminated, sometimes rapidly.

(C) To replace these sites, individual STVs can be captured (bottom line) and stabilized

(bottom box) at contact sites. t = time after contact, scale bars = 5m. (D) Histograms of

net changes in the densities of individual stable recruitment sites from one short imaging

sequence to the next. Stable sites were defined as puncta that appeared at a given site

throughout at least one short imaging sequence. The number of stable sites of recruitment

varied at the majority of contacts during both the rising and plateau phases, with a bias

toward puncta addition during the rising phase (left) but not the plateau phase (right).

69

Levels of synaptic vesicle proteins are highly labile at developing axo-dendritic

contacts

To determine if the same types of fluctuations in SV protein recruitment occur at

neuron-neuron synapses, we imaged regions of contact between axons expressing

synaptophysin-GFP and morphologically identified dendrites and somas expressing RFP.

Contacts were imaged for 12-24h at 7-10 DIV with the imaging protocol described above

(Figure 3.3A). Axo-dendritic and axo-somatic synapses exhibited many of the same

dynamics observed in axons contacting neuroligin-1-expressing HEK293 cells. Levels of

synaptophysin-GFP at individual sites of stable recruitment varied over time (Figure

3.3B,C). Synaptophysin-GFP protein levels at contact sites fluctuated 10.9% of the time

(n = 43) between short imaging sequences (over 1-2.5h) and 17.6% of the time (n = 49)

within short sequences (over 7.5min). Furthermore, individual recruitment sites were

established and dissolved over the course of hours (Figure 3.3B, C), with net changes in

the density of stable sites of recruitment occurring in 55.5% of imaging sequences

(Figure 3.3D; 31.1% additions and 24.4% losses; n = 45). These data indicate that the

dynamics of synaptic vesicle protein recruitment at axo-dendritic synapses and hemi-

synapses occurred over a similar time course. The similarities between recruitment

dynamics at spontaneous neuron-neuron synapses and neuroligin-1-induced sites are

especially striking since sites of recruitment may have been more mature in some neuron-

neuron synapses, and synapse formation between neurons may involve additional

signaling pathways. Together, the above data suggest that developing synapses undergo

extensive fluctuations in presynaptic protein levels, even during initial formation.

70

Figure 3.3: Synaptic vesicle protein recruitment to axo-dendritic contacts is similar

that at induced trans-synaptic signaling sites. (A) Image of an axon transfected with

synaptophysin-GFP (green) contacting somato-dendritic region from an adjacent cell

transfected with RFP (magenta). White box corresponds to contact region represented by

71

kymograph in (B). (B, C) Kymographs of STVs (green) in axons that contact RFP-

expressing dendrites from different neurons (red). Boxes, as seen in induced axo-

dendritic contacts, synaptophysin-GFP levels are variable at individual stable sites in

neuron-neuron contacts over the course of hours and minutes (asterisks). t = time after

imaging begins, scale bars (A) = 10m, (B, C) = 5m. (D) Histogram of net changes in

the densities of individual stable recruitment sites from one short imaging sequence to the

next for neuron-neuron contacts. The number of stable sites of recruitment fluctuated

between the majority of imaging sequences (corresponds to all non-zero bins), with a

small bias toward addition of stable sites of recruitment.

72

Synaptophysin is preferentially stabilized at clustered neurexin

Neuroligin induces synaptogenesis through its presynaptic binding partner

neurexin (5, 228). To ascertain the time course of neurexin clustering at contacts,

neuroligin-1-expressing HEK293 cells were added to neuronal cultures transfected with

neurexin-1β-GFP. Then contacts between neuroligin-1-expressing HEK293 cells and

axons expressing neurexin-1β-GFP were imaged immediately. Prior to contact, neurexin-

1β-GFP appeared dim and diffuse within the axon. Upon contact with neuroligin-1-

expressing HEK293 cells, neurexin-1β-GFP began to be recruited to sites of contact

within minutes, and showed strong recruitment within 1-2 hours (Figure 3.4A-C, n = 14

contacts), consistent with previous data (115). Because we have not found any neurexin-β

selective antibodies for immunofluorescence, we could not determine the extent to which

neurexin-1β-GFP was over-expressed in these experiments. However, similar

fluorescently tagged neurexin-1β-GFP has been used in a number of studies to investigate

the localization, trafficking and function of neurexin (229-232). Importantly, the

distribution pattern of α-neurexin is not significantly influenced by the level of neurexin

expressed (232). Therefore, although over-expression of neurexin in neurons increases

synapse density (230), it not expected to change the dynamics or distribution of neurexin.

Neuroligin-1 recruited neurexin-1β-GFP to the entire contact region, where

neurexin appeared in both punctate and diffuse forms, even within the same contacts

(Figure 3.4B). Recent work has shown that the synaptogenic effect of neuroligin depends

on the postsynaptic scaffolding protein S-SCAM, which works, in part, by clustering

neuroligin (216). To determine if post-synaptic clustering of neuroligin facilitates

presynaptic clustering of neurexin, HEK293 cells that co-expressed S-SCAM and

73

neuroligin-1 were dropped onto neurons that were transfected with neurexin-1β-GFP.

This led to enhanced clustering of neurexin-1β-GFP in axons that contacted transfected

HEK293 cells (Figure 3.4B), suggesting that the synaptogenic effects of S-SCAM-

induced neuroligin clustering might be due to enhanced clustering of presynaptic

neurexin. Indeed, artificial clustering of neurexin is sufficient to induce presynaptic

protein recruitment (124), and recent work suggests that neurexin clustering is critical for

the synaptogenic function of the neurexin-neuroligin interaction in Drosophila (125).

This raises the question of whether synaptic proteins are preferentially recruited to

punctate clusters of neurexin or whether they can be recruited to any site with neurexin

enrichment. To test this, neurons were co-transfected with neurexin-1β-GFP and

synaptophysin-RFP. Neuroligin-1-expressing HEK293 cells were then added to these

cultures, and contacts were imaged for up to 24h, as described above. Although there

was a combination of diffuse and punctate recruitment of neurexin-1β-GFP to contacts

(Figure 3.4A-C), stable sites of synaptophysin-RFP recruitment were almost exclusively

co-localized with neurexin-1β-GFP puncta (Figure 3.4C). Surprisingly, neurexin-1β-GFP

clustering and synaptophysin-RFP clustering occurred nearly simultaneously at most

shared sites (Figure 3.4D, n = 13/18 clusters). Neurexin clustered concurrently with SV

protein recruitment even when HEK293 cells expressed both neuroligin-1 and S-SCAM

to enhance neurexin clustering. These data support a model in which discrete neurexin

clustering is critical for stable, rapid, neuroligin-induced presynaptic protein recruitment

in mammalian cortical neurons.

74

Figure 3.4: Synaptic vesicle protein accumulation occurs at clustered neurexin and

correlates with levels of clustered neurexin. (A) Time-lapse images of an axon

transfected with neurexin-1β-GFP (green) contacting a neuroligin-1-expressingHEK293

cell (white outline). Neurexin is recruited to contact sites within minutes and enhanced at

these sites as imaging progresses. (B) Images of two axons that are transfected with

neurexin-1β-GFP (green) and contact cells expressing neuroligin-1 (left, white outline) or

neuroligin-1+S-SCAM (right, white outline). At neuroligin-1-only contacts, neurexin

appears both punctate (arrows) and diffuse (bars). Addition of S-SCAM enhanced

neurexin-1β-GFP clustering and reduced diffuse neurexin-1β-GFP labeling at contacts

75

(arrow). (C) Persistent sites of synaptophysin-RFP (magenta) recruitment appeared at

sites of punctate neurexin-1β-GFP (green) recruitment (arrows). t = time after contact

established. (D) Kymographs of neurexin-1β-GFP and synaptophysin-RFP, starting

immediately after contact with a neuroligin-1-expressing cell was established. Neurexin

and synaptophysin clustered simultaneously at the same sites in the axon (arrows). (E) At

an individual recruitment site, synaptophysin-RFP and neurexin-1β-GFP levels were

positively (r = 0.98) and significantly (p = 2.0x10-5

) correlated over several hours of

imaging. This correlation was significant for 11 of 14 co-recruitment sites that were

present for at least 5h. (F) When combining data from all recruitment sites in the same

axon, synaptophysin and neurexin integrated densities were also correlated (r = 0.91, p =

6.0x10-20

). Similar correlation was observed in 5 of 7 axons. (G) When the data from all

axons were pooled, neurexin and synaptophysin were positively and significantly

correlated (p = 1.0X10-74

, r = 0.82). (A-C) scale bars = 10m, (D) scale bar = 5m, (E-G)

correlations made from non-transformed data and plotted in log-log form for display

purposes.

76

Actin polymerization is occasionally co-localized with neurexin clustering

While the neuroligin-neurexin interaction by itself is capable of generating

synaptogenic signaling, other signaling pathways can interact with and enhance

neuroligin-neurexin-induced synapse formation. Specifically, n-cadherin can act in

concert with neuroligin to enhance synaptic vesicle recruitment to developing presynaptic

terminals (216, 233). N-cadherin can mediate its intracellular signaling through actin, and

actin is known to play a role in the recruitment of presynaptic components to the

developing synapse (234-236). Therefore, to determine whether actin plays a role in

mediating the initial recruitment of presynaptic components to synapses, the time-course

of actin polymerization to novel sites of neurexin signaling was determined. Specifically,

HEK293 cells expressing neuroligin were dropped onto 6-8 DIV neuronal cultures that

co-expressed neurexin-GFP and Lifeact-mcherry, which is a fluorescently-tagged small

peptide that binds endogenous F-actin (237, 238). HEK293 cells that contacted neurexin-

GFP and Lifeact-mcherry-expressing axons were then imaged immediately via live time-

lapse confocal microscopy at a frame rate of one frame per minute. Neurexin recruitment,

including neurexin clustering, was observed at multiple contact sites after 60 minutes.

Approximately half of the contact sites (n = 4/9 contacts in 4 separate axons) displayed F-

actin clustering 60 min. after the initiation of neurexin recruitment (data not shown).

The synaptic scaffolding molecule S-SCAM physically links n-cadherin with

neuroligin via intracellular interactions on the postsynaptic side of the synapse (216).

This linkage induces enhanced clustering of neurexin and enhanced synapse formation

(216)(Figure 3.4B). Since n-cadherin binding can mediate downstream actin

polymerization (235), we hypothesized that postsynaptically localizing n-cadherin with

77

neuroligin might enhance actin polymerization presynaptically, particularly at sites of

neurexin clustering. Therefore, HEK293 cells expressing neuroligin + S-SCAM were

dropped onto 7-8 DIV axons that co-expressed neurexin-GFP and Lifeact-mcherry and

immediately imaged. Importantly, n-cadherin is endogenously expressed in HEK293

cells (239). Neurexin clustering was enhanced at these contact sites, compared to contacts

with HEK293 cells that did not express S-SCAM. However, co-localized actin

polymerization and neurexin clustering was observed in only 33% of contacts (data not

shown, n = 4/12 contacts in 3 separate axons).

While it is tempting to infer that actin is not typically involved in the initial

recruitment of synaptophysin, these experiments were performed in a limited number of

neurons. In previous experiments, simultaneous co-clustering of neurexin and

synaptophysin were not always observed (n = 5/18 contacts), suggesting that specific

neurons or even specific sites within the same neuron are more or potentially less capable

of this rapid co-clustering. Follow-up experiments involving additional neurons and

HEK293 cell/axon contacts are needed to confirm these results both with and without S-

SCAM as a postsynaptic organizer.

Levels of recruited synaptophysin and neurexin are strongly correlated at

developing presynaptic terminals

Our observation that SV proteins and neurexin selectively and simultaneously co-

cluster raises the question of whether synaptophysin recruitment and neurexin clustering

are coordinated. To test this, we first determined whether synaptophysin and neurexin

levels were correlated at individual sites of recruitment. Indeed, synaptophysin levels

78

were positively and significantly correlated with neurexin levels at individual co-clusters

that were present for at least five consecutive short sequences of imaging (Figure 3.4E, p

< 0.05 at 11/14 sites from 7 axons, r = 0.37-0.98). Neurexin-1β-GFP and synaptophysin-

RFP levels were also significantly correlated when all clusters from individual axons

were pooled (Figure 3.4F, p < 0.05 in 5/7 axons, r =0.52-0.96) or when all axons were

combined (Figure 3.4G, n = 307 sites, p = 1.0X10-74

, r = 0.82). Therefore, recruitment of

SV proteins and clustered neurexin are coordinated during presynaptic terminal

assembly.

Next, we asked whether the long-term dynamics of synaptophysin and neurexin

recruitment were similar. To test this, 7-10 DIV neurons were transfected with either

synaptophysin-GFP and neurexin-tdTomato or the near-infrared fluorescent protein

RFP670 (Figure 3.5) (240). Contacts between axons that expressed both synaptophysin-

GFP and neurexin-tdTomato and dendrites that expressed RFP670 were then imaged over

the course of 12-25h. Like synaptophysin, levels of neurexin fluctuated significantly over

the course of long-term imaging at both spontaneously-formed axo-dendritic and axo-

somatic neuron-neuron contacts (Figure 3.5A,B) as well as contacts formed with

neuroligin-expressing HEK293 cells (Figure 3.5C). In addition, many neurexin-1β-GFP

clusters formed and disappeared over the course of hours (Figure 3.4C and 3.5),

resembling the formation and dissolution of stable sites of synaptophysin recruitment

(Figure 3.2 and 3.5).

79

Figure 3.5: Neurexin and synaptophysin levels fluctuate over the course of minutes

at axo-dendritic contact sites and neuroligin induced developing synapses. (A) Image

of an axon co-expressing synaptophysin-GFP (green) and neurexin-tdTomato (red) and

contacting the somato-dendritic region of a neuron expressing RFP670 (blue). White box

corresponds to contact region represented by kymographs in (B). (B) Individual

kymographs of an axon expressing synaptophysin-GFP (top) and neurexin-tdTomato

(middle) that contacts a dendrite expressing RFP670 (bottom). Levels of both neurexin

and synaptophysin fluctuate over the course of minutes to hours at sites that contact

somato-dendritic regions of the adjacent neuron (arrows, top), including the formation

(white box) and disappearance (white arrowhead) of co-clusters. (C) Kymographs of an

axon expressing neurexin-GFP (left) and synaptophysin-RFP (right) contacting a

neuroligin-expressing HEK293 cell. The amount of neurexin and synaptophysin at co-

clustered sites (arrow, bottom) fluctuates extensively over hours and even within minutes

(white arrowheads). Scale bars (A,B) = 10m, scale bar (C) = 5m.

80

Trans-synaptic signaling recruits synaptic proteins without substantially altering

their transport

It is not known how STVs disengage from their microtubule tracks and become

incorporated into synapses. Various signals that affect synaptogenesis also affect STV

transport (51), and it has been suggested that neuroligin-neurexin signaling might recruit

synaptic proteins by altering synaptic protein transport (241). Therefore, we next tested

the hypothesis that trans-synaptic signaling recruits presynaptic proteins by altering STV

transport. To test this hypothesis, synaptophysin-GFP was imaged every 10s for 7.5 min

at 1-4 hours after contact was established between axons and neuroligin-1-expressing

HEK293 cells (Figure 3.6A). Because STV transport is characterized by bursts of

movement interrupted by pauses (47, 48, 51, 53, 58, 87, 210), we measured how fast

STVs moved and how long they paused in the presence and absence of neuroligin-1. We

found that STV pause duration increased at contacts with neuroligin-1-expressing cells

when compared to contacts with HcRed-expressing control cells (Figure 3.6B; neuroligin

= 145.3s, +/- 20.8, n = 40 pauses; no neuroligin = 85.7s, +/- 15.5, p = 0.0217, n = 42

pauses). However, there was no neuroligin-1-dependent change in instantaneous velocity

(Figure 3.6B; neuroligin = 0.229m/s, +/- 0.023, n = 40 movements; no neuroligin =

0.196m/s, +/- 0.015, n = 56 movements, p = 0.391).

To investigate active zone protein transport, we looked at Piccolo-Bassoon

transport vesicles (PTVs), which deliver active zone proteins to developing presynaptic

terminals (49, 59-62, 73, 74, 81, 83, 87, 115, 118). STVs and PTVs can be transported

together in the axon (49, 87, 96). Therefore, neuroligin-neurexin interactions could alter

PTV transport, which in turn could recruit STVs. Similar to STVs, neuroligin-1 had no

81

effect on the instantaneous velocity of PTVs (Figure 3.6C; neuroligin = 0.140m/s, +/-

0.008, n = 22 movements; no neuroligin = 0.158m/s, +/- 0.009, n = 59 movements, p =

0.473). Unlike STVs, contact with a neuroligin-1-expressing cell had no effect on the

pause duration of PTVs (Figure 3.6C; neuroligin = 173.1s, +/- 24.9s, n = 32 pauses; no

neuroligin = 145.5s, +/- 14.0s, n = 53 pauses, p = 0.634). Therefore, neuroligin-neurexin

signaling does not recruit STVs indirectly via regulation of PTV transport.

To determine if other forms of synaptogenic trans-synaptic signaling alter STV

movement, we also performed the live-imaging assay described above using HEK293

cells transfected with SynCAM-1, a trans-synaptic adhesion molecule that possesses

synaptogenic properties similar to neuroligin-1(129, 143, 156). Unlike neuroligin-

neurexin signaling, SynCAM-1-mediated trans-synaptic adhesion did not alter STV pause

duration (Figure 3.6D; SynCAM = 112.5s, +/- 14.0, n = 53 pauses; no SynCAM =

125.5s, +/- 13.4, n = 58 pauses, p = 0.332). However, SynCAM-1 signaling caused a

small increase in the instantaneous velocity of STVs (Figure 3.6D; SynCAM =

0.235m/s, +/- 0.021, n = 52 movements; no SynCAM = 0.183m/s, +/- 0.012, n = 58

movements, p = 0.030). Since the time-course of presynaptic protein recruitment to

SynCAM-1 adhesion sites has not been established, we also imaged STVs at 24h after

contact with SynCAM-1. After 24h, SynCAM-1-expressing cells had no effect on STV

movement (Figure 3.6E; pause duration: SynCAM = 140.3s, +/- 15.6, n = 62 pauses; no

SynCAM = 116.5s, +/- 14.4, p = 0.197, n = 71 pauses, instantaneous velocity: SynCAM

= 0.202m/s, +/- 0.012, n = 71 movements; no SynCAM = 0.171m/s, +/- 0.007, p =

0.060, n = 86 movements).

82

Neuroligin and SynCAM interact with different presynaptic adhesion molecules

(5, 136, 242). Therefore, it might be expected that the effects of neuroligin and SynCAM

would be additive or synergistic. However, contact with HEK293 cells that expressed

both neuroligin-1 and SynCAM-1 had no effect on the movement of STVs (Figure 3.6F;

pause duration: neuroligin/SynCAM = 124.6s, +/- 12.7, n = 61 pauses; no

neuroligin/SynCAM = 116.7s, +/- 14.3, p = 0.482, n = 51 pauses; instantaneous velocity:

neuroligin/SynCAM = 0.196m/s, +/- 0.012, n = 63 movements; no neuroligin/SynCAM

= 0.175m/s, +/- 0.008, p = 0.187, n = 73 movements).

83

Figure 3.6: Trans-synaptic signaling has little effect on synaptic vesicle or active

zone protein transport. (A) Time-lapse images of a neuroligin-1 + HcRed-expressing

HEK293 cell (magenta) contacting an axon expressing synaptophysin-GFP (green).

Arrows, position of individual STV in each frame. Scale bar = 10m. (B) Contact with

neuroligin-1-expressing cells increases the mean pause duration but not instantaneous

velocity of STVs inside the contact region. (C) For PTVs, neither the mean pause

duration nor the instantaneous velocity was affected by neuroligin-1-induced trans-

synaptic signaling. (D) Contact with SynCAM-1-expressing cells for 1-4h did not affect

84

STV pause duration but increased their instantaneous velocity. (E) Contact with

SynCAM-1 for 24h had no effect on STV pause duration or instantaneous velocity. (F)

Contact with both neuroligin-1 and SynCAM-1 had no effect on STV pause duration or

instantaneous velocity. *, p < 0.05; t = time after imaging begins; error bars = S.E.M.

85

In addition to comparing the transport of STVs and PTVs in axonal regions that

contact HEK293 that either did or did not express synaptogenic adhesion molecules, we

also compared axonal transport inside and outside of contact regions within the same

axon. Each combination of transport vesicle and non-neuronal cell treatment described

above was tested. No significant (all p > 0.05) difference in average pause duration or

instantaneous velocity was observed when comparing transport vesicles inside of contact

regions vs. those outside of contact regions (data not shown), with one exception. The

average pause duration of STVs was significantly increased inside axonal regions that

contacted SynCAM-expressing cells for 24h (outside contact = 105.6s +/- 6.1, n = 339

pauses; inside contact = 140.3s +/- 15.6, n = 62 pauses).

Together, our data suggest that while trans-synaptic signaling might have some

effect on STV movement, regulation of transport is not the main mechanism through

which trans-synaptic adhesion mediates synaptic protein recruitment. It is unclear

whether the distinct effects of neuroligin and SynCAM on STV trafficking represent

mechanistic differences in how recruitment of STVs occurs down-stream of presynaptic

neurexin and SynCAM. Since neuroligin plus SynCAM did not result in the same

changes as either neuroligin or SynCAM, it is possible that either (i) the observed

changes are not essential for STV recruitment, or (ii) when accumulated at the same sites,

presynaptic neurexin and SynCAM compete with each other for some factor that is

necessary for the observed effects on STV trafficking, essentially dampening their

distinct effects.

Synaptic vesicle and active zone protein transport vesicles are not attracted to sites

of trans-synaptic signaling

86

Synaptogenic adhesion could recruit presynaptic machinery by actively attracting

synaptic proteins to sites of trans-synaptic signaling. To test this hypothesis, we

determined whether STV movements are biased towards sites of neuroligin-1 contact

(Figure 3.7). First, we quantified the net distance and direction moved for individual

STVs. Movements toward the contact were assigned positive values, while movements

away from the contact were negative. This analysis indicated that neuroligin-1 signaling

did not bias net STV movement toward neuroligin-1 (Figure 3.7B; net movement = -

0.078 +/- 0.857m, n = 67 STVs), similar to contacts with HcRed-expressing HEK293

cells (Figure 3.7C; net movement = -1.39 +/- 1.22m, n = 44 STVs; p = 0.60). Since

trans-synaptic signaling might only attract vesicles near the contact, we then restricted the

analysis to STVs with initial positions adjacent to the contact, within a region equal to the

contact in length. Again, neuroligin-neurexin signaling did not bias STV transport

(Figure 3.7D; Nlgn = 0.000 +/-0.888m, n = 17 STVs; HcRed = -3.08 +/- 1.69m, n = 19

STVs; p = 0.23). Moreover, the distributions of the net movements were not altered by

neuroligin-neurexin signaling (p > 0.15, Kolmogorov-Smirnov). In addition, neuroligin-1

did not change the fraction of STVs moving toward contacts (Figure 3.7E; # STVs

moving toward/# STVs moving away: Nlgn = 0.89, n = 66 STVs; HcRed = 0.83, n = 44

STVs). Similarly, STVs adjacent to contacts were not attracted to contacts (Figure 3.7F;

Nlgn = 0.89, n = 17 STVs; HcRed = 0.46, n = 19 STVs).

STVs also displayed no bias in movement towards HEK293 cells expressing

SynCAM-1- or SynCAM-1+neuroligin-1. SynCAM-1did not alter the net distance and

direction moved by all imaged STVs (p = 0.33, n = 136 and 88 STVs for SynCAM and

HcRed, respectively) or STVs adjacent to contacts (p = 0.23, n = 39 and 28 STVs for

87

SynCAM and HcRed). The same was true when contacts expressed both SynCAM-1 and

neuroligin-1 (all STVs: p =0.70, n = 136 and 69 STVs for SynCAM+neuroligin and

HcRed, respectively; near contacts: p = 0.55, n = 39 and 18 STVs for

SynCAM+neuroligin and HcRed, respectively). Likewise, STV movement was not

biased toward contacts with SynCAM-1after 24h of contact (p = 0.13, n = 39 and 36

STVs for SynCAM and HcRed, respectively). Again, the data were not significantly

different when the distributions were compared using the Kolmogorov-Smirnov test (p >

0.11). Furthermore, the ratio of STVs moving toward vs. away from contact regions (for

SynCAM, SynCAM+neuroligin, or 24hr SynCAM) was less than or equal to 1.13,

indicating a lack of attraction to sites of trans-synaptic adhesion. Taken together, the

above data demonstrate that trans-synaptic signaling does not actively recruit STVs from

outside regions of contact.

Although STVs were not actively attracted to contact sites, it remained possible

that trans-synaptic signaling actively recruits active zone precursors to sites of contact,

followed by passive capturing of STVs. To test this, we determined whether movements

of PTVs were biased toward trans-synaptic adhesion sites. Similar to STVs, PTVs

showed no bias in movement towards neuroligin-1 contact sites (Figure 3.7G), both when

including all PTVs in the analysis (Nlgn = -0.35+/-0.56m, n = 79 PTVs; HcRed =

1.38+/-0.68, n = 69 PTVs; p = 0.071) and when limiting the analysis to PTVs with initial

positions adjacent to contacts (Nlgn = -1.34+/-1.12m, n = 26 PTVs; HcRed = 0.72+/-

0.86, n = 28 PTVs; p = 0.046). When the distributions were compared using the

Kolmogorov-Smirnov test, no differences were seen (p > 0.084). Collectively, our data

88

indicate that trans-synaptic adhesion signaling does not attract synaptic vesicle or active

zone matrix proteins to developing synapses.

89

Figure 3.7: Trans-synaptic signaling does not attract synaptic vesicle or active zone

proteins. (A) Schematics of HEK293 cell contact with an axon. Analysis was done on all

STVs with initial positions either outside the contact area (gray area, left, center) or

directly adjacent to the contact region (gray area, right). (B-D) Histograms of the net

movement of STVs towards (positive) or away from (negative) contacts with HEK293

cells expressing either neuroligin-1 + HcRed (B, D) or HcRed only (C). STV movement

was not biased towards sites of neurexin-neuroligin signaling, for either all STVs or only

STVs adjacent to the contact. (E, F) Scatter plot of the number of STVs moving towards

vs. away from contacts with cells expressing neuroligin-1 + HcRed (blue circles) or

90

HcRed only (red squares). Each small point represents the ratio for one axon, with larger

points representing additional axons with the identical ratio. These ratios were

approximately the same for axons that contact cells expressing neuroligin-1 or HcRed

(black line = unity ratio line). (G) Similar to STVs, PTVs were not attracted to sites of

neuroligin-1 contact. (B-D, G) Red arrows = mean.

91

Initial arrival of STVs and PTVs to a developing synapse is likely uncoordinated

Due to the variety of synaptic components transported by STVs and PTVs, it is

possible one type of transport vesicle is initially recruited to a developing synapse and

subsequently traps the other type of transport vesicle and/or additional transport vesicles

of the same type. If this is the case, then one of these transport vesicles would always

arrive at a developing synapse before the other. Therefore, to determine whether STVs or

PTVs are recruited to novel presynaptic terminals first, 6-8 DIV neurons were transfected

with either RFP670 or bassoon-GFP + synaptophysin-RFP. Axons that co-expressed

bassoon-GFP and synaptophysin-RFP and made contact with RFP670-expressing

dendrites were then imaged at a rate of 1 frame per 90s for up to 24 hours. During this

time, newly stabilized presynaptic terminals were formed along the axo-dendritic contact.

Synapse stabilization was defined as the co-recruitment of synaptophysin-RFP and

bassoon-GFP to the same site in the axon where both components remained at that site

continuously for at least one hour.

Overall, 51 newly stabilized presynaptic terminals were observed at 11 unique

neuron-neuron contacts. Upon visual examination of the images, it appeared that

synaptophysin-GFP arrived at the synapse first 31% of the time (n = 16/51 stabilized

synapses), bassoon-GFP arrived at the synapse first 33% of the time (n =17/51 stabilized

synapses), and simultaneous recruitment of synaptophysin-GFP and bassoon-GFP

occurred 35% of the time (n = 18/51 stabilized synapses, data not shown). This implies

that there is no defined order for when STVs or PTVs arrive at the synapse. In addition,

the high proportion of simultaneous arrivals suggests that when synaptogenic signaling

92

occurs, both STVs and PTVs can be recruited to the same site rapidly, at least within 90

seconds of each other, as that was the length of time between imaging frames.

There are caveats to this analysis, however. The frame rate of 1 frame per 90s was

utilized to maximize temporal resolution while maintaining cell health over long periods

of time. Long-duration imaging was necessary because novel formation of presynaptic

terminals is relatively rare on existing axo-dendritic contacts. It is possible that one type

of transport vesicle is always stabilized first and then rapidly recruits the other type of

vesicle into the synapse. In addition, although there were many instances of STVs or

PTVs arriving to a new synaptic site first, it was often difficult to distinguish when one of

these components was initially stabilized at a specific site because both STVs and PTVs

appeared to pause at these sites before becoming more “permanently” fixed to the site.

This poses a challenge for analysis, as it would be impossible to distinguish between a

transport vesicle that pauses at the same site every 90s during the imaging of each frame

and a truly stabilized synaptic component. Therefore, higher temporal resolution is

needed to distinguish between these two phenomenon. Ongoing experiments will utilize

the synapse induction and long-term imaging protocol described in Figures 3.1, 3.2, and

3.3 to address this issue.

93

Materials and Methods

All studies were conducted with an approved protocol from the Case Western

Reserve University Institutional Animal Care and Use Committee, in compliance with the

National Institutes of Health guidelines for the care and use of experimental animals.

Neuronal culture and transfection

Primary neuronal cultures were prepared from rat or mouse cortices on 18mm

glass coverslips as described previously (50, 51, 87, 207) and maintained in Neurobasal

A media with B27 Supplement (Invitrogen, Carlsbad, CA, USA). At 6-9 days in vitro

(DIV) and 24-48h prior to imaging, neurons were transfected using Lipofectamine 2000,

essentially according to the manufacturer’s instructions (Invitrogen). DNA was used at

1μg per 18mm coverslip with the exception of GFP-bassoon, which was transfected at

2μg per 18mm coverslip due to its large size. This amount of DNA results in efficient

labeling of transport vesicles without significant over-expression within axons (50, 51,

53, 87). Even so, neurons were chosen for imaging that displayed moderate expression

levels of the transfected constructs to avoid any possible effects of over-expression. For

double transfections, the localization and movement of fluorescently-tagged proteins

appeared similar to single transfections. In addition, synaptic vesicle protein transport

vesicles (STVs) labeled with synaptophysin-RFP and synaptophysin-GFP were similar in

size, movement, and localization.

Synaptophysin-GFP, GFP-bassoon (GFP-Bsn 95-3938), synaptophysin-mRFP,

synaptophysin-mcherry, and neurexin-1β-GFP were generous gifts of Drs. Jane Sullivan

(University of Washington, Seattle), Thomas Dresbach (Georg August University,

94

Gottingen), Jurgen Klingauf (University of Muenster), Matthijs Verhage (Vrije

Universiteit, Amsterdam), and Camin Dean (The European Neuroscience Institute,

Gottingen). These constructs have been shown to be functional and properly localized

and have been used previously for studies of protein trafficking and localization (50, 61,

115, 211, 213, 243). Fluorescent proteins attached to bassoon and synaptophysin

effectively label PTVs/active zones and STVs/synaptic vesicles, respectively (48, 50, 53,

60, 61, 87, 115, 210, 211). RFP670 was purchased from Addgene [plasmid # 45457].

Neurexin-tdTomato was generated by seamlessly replacing the GFP sequence in the

neurexin-GFP construct with tdTomato via Gibson Assembly (New England

Biosystems).

HEK293 cell culture and transfection

HEK293 cells were cultured and maintained in DMEM supplemented with 10%

fetal calf serum and penicillin/streptomycin (Invitrogen, HyClone). Cells were

transfected 24-48h prior to dissociation and dropping onto neuronal cultures. A 3:2 or 2:1

ratio of trans-synaptic adhesion DNA to HcRed DNA (Clontech) was used to ensure co-

expression of both types of DNA constructs in the same HEK293 cell (51). To confirm

co-expression, HEK293 cells were transfected with SynCAM-GFP, HA-neuroligin-1, and

HcRed. After 48h, cells were fixed and labeled with anti-HA antibody to detect HA-

neuroligin-1. Over 85% of HEK293 cells transfected with HcRed also expressed

SynCAM-GFP and HA-neuroligin-1 (data not shown). HA-neuroligin-1 and HA-

SynCAM-1 were generous gifts of Drs. Peter Scheiffele (University of Basel) and Philip

Washbourne (University of Oregon, Eugene) and have previously been shown to recruit

synaptic proteins in the hemi-synapse assay (128, 244). Myc-tagged S-SCAM was

95

provided by the labs of Yutaka Hata (Tokyo Medical and Dental University, Tokyo,

Japan) and Yoshimi Takai (Kobe University, Kobe, Japan) (245) and purchased from

Addgene (plasmid # 40213245). pDEST/Lifeact-mcherry-N1 was provided by Dr. Robin

Shaw (Cedars-Sinai, Los Angeles, CA, USA) (238) and purchased from Addgene

[plasmid # 40908].

Imaging

Imaging was performed with a C1 Plus confocal system on a Nikon Eclipse Ti-E

microscope utilizing a 40x Nikon Plan Apo 0.95NA objective. Lasers used for excitation

were 488nm argon and 543nm and 633 nm helium-neon, while detection filters were

515/30nm bandpass for GFP and 590/50nm bandpass for HcRed/mcherry/mRFP. For

multi-color imaging, channels were imaged sequentially to avoid bleed-through. To avoid

focal plane drift, the Perfect Focus System (Nikon) was used.

Long-term imaging

Transfected HEK293 cells were dissociated with Hank’s-based, enzyme-free cell

dissociation buffer (Invitrogen). Cells were resuspended in neuronal media then 8 X 104

cells were dropped onto each coverslip of transfected neurons. Contacts between

transfected axons and transfected HEK293 cells were imaged, with axons identified by

their morphological characteristics (50, 51). Dual-color images were collected every 10s

for 7.5min, with scan times for both frames not exceeding 3.3s. This interval yields high

temporal resolution of STV movement with minimal phototoxicity (87). This imaging

protocol was repeated every hour for the next 4h then every 2.5h for up to 25h. To ensure

that cells remained healthy, imaging was performed in a custom made environmental

96

chamber held at 32°C. The pH was maintained at physiological levels by equilibration in

5% CO2 for 15-30min prior to sealing with vacuum grease (Dow Corning).

Short-term imaging

Synaptogenesis was induced as described above. After 1-4h, contacts between

transfected axons and transfected HEK293 cells were imaged every 10s for 7.5min with

constant perfusion of artificial cerebrospinal fluid (ACSF: 120mM NaCl, 3mM KCl,

2mM CaCl2, 30mM D-glucose, 20mM HEPES, and 0.2% sorbitol, pH 7.3) at 25°C. At

this temperature, STV movement is not significantly altered compared to STV movement

at physiological temperatures (51). In previous studies, the fastest velocities recorded for

STV movements were approximately 1μm/s (51, 53, 210), while maximal PTV velocities

were lower (60, 73). Vesicles moving at this maximal velocity would move 10μm

between frames during of imaging, a small fraction of the total axonal area typically

imaged. Therefore, the rapid imaging of STVs/PTVs performed here enabled reliable

identification and tracking of vesicles.

Neurexin-GFP recruitment

Axons expressing neurexin-GFP were imaged for at least 5min before contact

with HEK293 cells to identify neurexin-GFP that was previously clustered through

endogenous mechanisms. HEK293 cells transfected with HA-neuroligin-1 plus HcRed

were added to the imaged coverslip until at least one transfected HEK293 cell made

contact with the neurexin-expressing axons. This contact was then imaged for up to 2h at

a rate of 1 frame/min.

Analysis

97

STVs and PTVs were manually tracked using custom written macros for ImageJ

(NIH, Bethesda, MD, USA), while blind to the locations of the HEK293 cells. Analysis

was restricted to axons that did not move during imaging and where at least one vesicle

moved within the imaged area. To facilitate tracking, imaging was conducted on axons

with an intermediate to low density of STVs or PTVs. Movements were quantified by

importing the positional data into Matlab and subjecting them to custom written analysis

(Mathworks, Natick, MA, USA). Pausing was defined as a period of at least 10s where

the velocity was lower than 0.1μm/s, as previously described (51, 87). Vesicles that did

not move during imaging were not included in this analysis, and analysis was restricted to

vesicles that could be tracked throughout the imaging period. Therefore, very long

pauses and vesicles that moved at high velocities for long periods of time might be

underrepresented. Considering the scan speed, pixel density, and average size of

transport vesicles, a typical STV or PTV was imaged within 10-50ms.

Integrated density analysis was performed using custom-written macros in ImageJ

that determined the location and integrated density of each transport vesicle on the axon

for each image in a sequence. Contact regions were defined as axonal regions that

overlapped with fluorescence from a transfected HEK293 cell. ‘% Enrichment’ was

determined by the following equation:

% Enrichment = [(IntDenin - IntDenout)/ (IntDenin + IntDenout)]*100

Where IntDenin corresponded to the mean integrated density per m of axon length inside

the contact, and IntDenout corresponded to the mean integrated density per m outside the

contact for the same axon. Contacts were considered to display recruitment when the %

98

enrichment was (i) positive for at least 70% of time points 4 hours after contact was

initiated and (ii) greater than 10% for at least half of all time points 4+ hours after contact

initiation. Data are shown as the mean +/- standard error when applicable. Significance

was determined via Wilcoxon rank sum test unless otherwise indicated. For correlation

data, r = Pearson’s linear correlation coefficient and p-values were determined using

Student’s t-test.

Acknowledgements

We thank Michael Sceniak for helpful discussions and suggestions on the project. We

also thank Valentina Savchenko, Teresa Evans, Ruth Siegel, Lynn Landmesser, and

Brian McDermott for helpful reading of the manuscript. In addition, we thank Louie

Zhou and Jeesoo Kim for technical assistance, Ryan Yavorsky for assistance with

analysis and Corbett Berry for assistance with ImageJ programming. This work was

funded by National Institute of Mental Health Grant R01MH096908 (SLS), Simons

Foundation Autism Research Initiative (SLS), PhRMA Foundation Research Starter

Award (SLS) and the NIH Predoctoral Training Program in Molecular Therapeutics at

Case Western Reserve University (LADB).

99

Chapter 4: Imaging presynaptic dynamics during mouse cortical development via

two-photon confocal microscopy

Teresa A. Evans1,2*

, Luke A.D. Bury1*

and Shasta L. Sabo1,2

Departments of 1Pharmacology and

2Neuroscience, Case Western Reserve University

School of Medicine, Cleveland, Ohio.

*For individual contributions, please see Materials and Methods

100

Introduction

The previous experiments described in Chapters 2 and 3 have determined how the

transport of presynaptic components is coordinated prior to synapse formation and how

recruitment of these components progresses in response to induced synaptogenic

signaling. These experiments were done exclusively in living cultured cortical neurons,

which allowed for easy manipulation and high spatial and temporal resolution when

performing live, time-lapse confocal imaging. While this in vitro system offers many

benefits, it lacks key characteristics that are present in the intact brain. Specifically,

normal connectivity that develops between specific brain regions is not recapitulated in

neuronal cultures. In addition, many developing cortical circuits often require

environmental stimuli to refine their connectivity, which is also lacking in culture.

Understanding how the dynamic processes described above proceed within the

intact mammalian brain would therefore yield critical insight into how cortical circuits

are shaped. To achieve this insight, Cre-induced synaptophysin-tdtomato was

exogenously expressed in the mouse brain by viral injection of Cre-GFP. Cranial

windows were applied to these mice and synaptophysin-tdtomato was imaged within the

brain via 2-photon microscopy over the course of minutes. To detect whether presynaptic

dynamics change over the course of development, multiple imaging time-courses were

taken from the same animals on separate dates ranging from postnatal day 13 (P13) to

P22. These experiments are the first to detect fluorescently-tagged presynaptic proteins

within axons of the intact mammalian brain and add insight into synaptic dynamics that

occur during development.

101

Results

To analyze presynaptic dynamics during development in vivo, presynaptic

terminals must be both labeled and detected within the brain. In order to fluorescently

label presynaptic terminals, “stop-flox” synaptophysin-tdtomato transgenic mice were

obtained. These mice possess an allele that contains a loxP-flanked stop codon that is

upstream of a synaptophysin-tdtomato fusion gene (Figure 4.1A). This stop codon

normally inhibits synaptophysin-tdtomato expression. However, with Cre expression, the

stop codon is removed and synaptophysin-tdtomato is expressed (Figure 4.1A). To

induce synaptophysin-tdtomato expression in the brain, an adeno-associated virus (AAV)

encoding Cre-GFP was injected via Hamilton syringe into the right hemisphere of

neonatal pups aged P2-P4. This limited expression to cell bodies located in the right

hemisphere of the cortex (Figure 4.1B). In addition, the expression of Cre-GFP was

driven by the synapsin promoter, which restricted expression to neurons.

In order to detect Cre-induced synaptophysin-tdtomato expression, cranial

windows were applied at postnatal day 10 to mice which had previously been injected

with Cre-GFP AAV. Windows were applied over the left hemisphere of the cortex in

order to image synaptophysin-tdtomato expression on the contralateral side of the

injection (Figure 4.1C). This restricted the observable synaptophysin-tdtomato signal to

callosally projecting axons in the left hemisphere whose somas were located in the right

hemisphere. This imaging strategy provided a number of distinct advantages compared to

imaging the ipsilateral hemisphere. First, since only axons cross the midline and project

to the contralateral side of the brain, the synaptophysin-tdtomato puncta that were

observed were strictly axonal and not mislocalized to dendrites. In addition, only

102

excitatory neurons produce callosal projections, thereby limiting the synapses that are

detected to a specific population of cortical neurons. This neuronal population is

clinically relevant as well, as defects in callosal development have been linked to a

number of neurological disorders (246, 247).

103

Figure 4.1: Synaptophysin-tdtomato expression in transgenic knock-in mouse. (A)

Schematic for synaptophysin-tdtomato expression in stop-flox synaptophysin-tdtomato

mice. Injection of AAV encoding Cre-GFP downstream of the synapsin promoter induces

Cre expression. In stop-flox synaptophysin tdtomato mice, this Cre expression drives

recombination of the loxP sites, thereby removing the stop codon and inducing

synaptophysin-tdtomato expression in neurons. (B) Image of brain expressing Cre-GFP

(green) in the right hemisphere (white circle). Cranial window application and

subsequent imaging occurs on the contralateral side (white square, arrow). (C) Example

of single image taken in living mouse via two-photon confocal imaging of

synaptophysin-tdtomato (yellow) through a cranial window. Scale bar = 20m.

104

To determine if synaptophysin-tdtomato effectively localized to presynaptic

terminals in vivo, stop-flox synaptophysin-tdtomato mice that were injected with Cre-

GFP were perfused, sectioned, and labeled with antibodies against the presynaptic protein

synapsin. As highlighted by the white circles in Figure 4.2A, synaptophysin-tdtomato

colocalized frequently with synapsin labeling. To further confirm the synaptic

localization of synaptophysin-tdtomato, stop-flox synaptophysin-tdtomato mice were

cross-bred with Thy-1-YFP mice which express YFP in a sparse population of layer V

pyramidal cells. After injection with the Cre-GFP virus, the brains of these mice were

perfused and sectioned. As shown in Figure 4.2B, synaptophysin-tdtomato puncta

frequently and closely apposed the apical dendrite of YFP+ cells, suggesting a synaptic

localization. It should be noted that, although these data suggest a synaptic localization

for many, if not most synaptophysin-tdtomato puncta, it does not rule out the possibility

of some synaptophysin-tdtomato signal being non-synaptic and/or part of transport

vesicles within the axon.

105

Figure 4.2: Synaptophysin-tdtomato is localized to presynaptic terminals. (A)

Synaptophysin-tdtomato (magenta) in the cortex co-localizes with antibody-labeled

synapsin (green). (B, left image) Multiple synaptophysin-tdtomato puncta (magenta) are

seen closely apposed to the apical dendrite of a YFP-filled cell (green). White circles and

square highlight co-localization points. (B, right images) Z-projections of the area

highlighted by the white square further indicate that synaptophysin-tdtomato and

dendrites are closely apposed in three-dimensional space. Scale bar for left image in (B)

= 10m, scale bar for right images in (B) = 1m.

106

In order to monitor presynaptic dynamics within the intact brain, synaptophysin-

tdtomato was imaged through cranial windows in living mice via two-photon time-lapse

imaging. The frame rate of all time-lapse movies was one frame per thirty seconds, and

images were collected for up to one hour. In between each imaging frame, 30 seconds of

rest without imaging was performed to prevent overheating of the brain tissue. In some

instances, multiple time-lapse images of the same animal over the course of days were

obtained. This was not possible in all cases, however, as bone growth under the cranial

window occasionally obscured synaptophysin-tdtomato detection. Upon imaging, it was

determined that the majority of synaptophysin-tdtomato was detected in superficial areas

near the pial membrane, likely in layer I of the cortex. Therefore, our analysis focused on

these synapses which were consistently and strongly labeled across multiple animals over

multiple days of imaging. In order to determine whether the dynamics of presynaptic

terminals in this cortical layer change during development, imaging was done in animals

ranging in age from P13-P22, a time period in which critical synaptic development takes

place (248, 249). Movies were then binned according to specific developmental age.

Both splitting and merging of individual synaptophysin-tdtomato puncta were

observed at all points in development (Figure 4.3A,B). Both the percent of puncta that

split per minute of imaging and the percent of puncta that merged per minute of imaging

increased during the P16-P19 time period (compared to P13-P15) before becoming

reduced back to near original levels during the P20-P22 time period (for all percentages,

value = (# split puncta in time-lapse series/ total puncta in time-lapse series / length of

time-lapse series): for P13-P15, % puncta split per minute = 0.038, +/- 0.013, n = 6 time-

lapse series; % puncta merged per minute = 0.067, +/- 0.027, n = 6 time-lapse series; for

107

P16-P19, % puncta split per minute = 0.14, +/- 0.032, n = 14 time-lapse series; % puncta

merged per minute = 0.18, +/- 0.034, n = 14 time-lapse series; for P20-P22, % puncta

split per minute = 0.066, +/- 0.018, n = 5 time-lapse series; % puncta merged per minute

= 0.11, +/- 0.042, n = 5 time-lapse series). The rates for both processes were quite similar

at each time period analyzed, suggesting that these processes likely do not contribute to

any overall changes in synapse density that occur during development. However, splitting

and merging of presynaptic components likely plays at least some role in refining circuits

over time. The extent to which splitting and merging of existing synaptophysin-tdtomato

structures contributes to this refining process, as well as the mechanisms responsible for

these phenomena remain to be determined.

It should be noted that spatial resolution is limited in the axial (Z) dimension due

to the point spread function of light detected through the objective. This could lead to

false detection of merging events, if one synaptophysin-tdtomato puncta moves directly

above or below another puncta and remains in close proximity to that puncta. It might

also lead to lower detection of splitting events, if a single puncta splits into two distinct

puncta that separate in the axial direction and do not move away from each other.

Resolution of the lateral (X, Y) dimensions is better than that of the axial dimension.

However, small distinct puncta that are in close lateral proximity to each other might be

detected as a single larger puncta or larger puncta that split but do not move away from

each other would not be detected. Again, this would lead to an overrepresentation of

merging events and an underrepresentation of splitting events in our data.

108

Figure 4.3: Synaptophysin-tdtomato puncta both split and merge in vivo. (A) An

example of two synaptophysin-tdtomato puncta (red) merging over the course of minutes

within the intact mouse brain. (B) The percent of synaptophysin-tdtomato puncta that

merge and split is similar over the same developmental time-course, suggesting that these

processes do not account for major changes in synapse number at this developmental

stage. However, circuit refinement is a likely consequence of these dynamics.

109

In addition to splitting and merging, a small population of synaptophysin-

tdtomato puncta was quite mobile within the cortex. These puncta were manually tracked

via Imaris software and subjected to analysis via custom-written MATLAB programming

(Figure 4.4A). Both instantaneous velocity and puncta displacement were measured for

mobile synaptophysin-tdtomato at each developmental time point. At the earliest

timepoint (P13-P15), the average instantaneous velocity was 0.141m/s, +/- 6.2x10-3

mm/s (Figure 4.4B, n = 35 movements), while it was slightly but significantly lower

during the two subsequent time points, indicating a reduction of rapid movement for the

most mobile puncta (Figure 4.4B, for P16-P19, inst. vel. = 0.129m/s, +/- 3.3x10-3

, n =

97 movements, p = 0.040 (compared to P13-15); for P20+, inst. vel. = 0.124m/s, +/-

7.9x10-3

n = 13 movements, p = 0.044 (compared to P13-15)). The average total distance

moved per puncta (normalized for length of the individual time-lapse image) was also

decreased with developmental age as this value was significantly lower at the latest time

point (Figure 4.4C, at P20+ distance moved = 0.15 m/puncta/min, +/-

0.045m/puncta/min, n = 8 puncta), compared to the two earlier time periods (Figure

4.4C, P13-15 distance moved = 0.70 m/puncta/min, +/- 0.13m/puncta/min, n = 16

puncta, p = 0.0016 (compared to P20+); P16-19 distance moved = 0.68m/puncta/min,

+/- 0.063m/puncta/min, n = 63 puncta, p = 4.8x10-5

(compared to P20+)). These data

taken together suggest that presynaptic terminals in the developing cortex become less

dynamic as time progresses.

110

Figure 4.4: Instantaneous velocity and mean displacement decrease as development

progresses. (A) Series of time-lapse images showing movement of two different

synaptophysin-tdtomato puncta over time (white arrow, red arrowhead). (B)

Instantaneous velocity of moving puncta decreases slightly as development progresses.

(C) Normalized distance moved per minute also decreased at later developmental time

points. Scale bar = 10m, * = p-val < 0.05, ** = p-val < 0.005

111

Materials and Methods

Contributions

Time-lapse two-photon imaging was performed by TAE while tracking analysis with

Imaris software and MATLAB programming was performed by LADB.

Virus injection

LSL-synaptophysin-tdtomato homozygous mice were obtained from Jackson

Laboratories (stock #: 012570 (tdtom), Bar Harbor, ME, USA) and bred to obtain pups

with the flox-stop synaptophysin-tdtomato gene. For some experiments, these mice were

cross-bred with Thy-1-YFP mice (Jackson Laboratories, Bar Harbor, ME, USA). At

postnatal day 2-4, pups were anesthetized via hypothermia, after which AAV1-hSyn1-

Cre-IRES-GFP virus was injected into the right hemisphere of the cortex via Hamilton

syringe. Virus was created by the Vector Core at the University of Pennsylvania

(Philadelphia, PA, USA). Pups were then warmed by hand before being placed back into

their original cage.

Cranial window

At P10, pups that previously had undergone viral injection were anesthetized via

inhalation of approximately 1-3% isoflurane solution vaporized in pure O2 dispensed at a

rate of 1.5L/min. Pups were then given a subcutaneous injection of dexamethasone

(0.1mg/kg) and carprofen (5mg/kg) to prevent post-surgery edema and pain, respectively.

The head was then stabilized via stereotaxic instrument (Stoelting Co., Wood Dale, IL,

USA) before removing the skin on top of the head to expose the skull. A surgical

microdrill (Fine Science Tools, Foster City, CA, USA) was then utilized to gently cut

112

around and remove a section of the skull over the left hemisphere to expose the brain.

During drilling and after removal of skull piece, the surgical region was constantly

submerged in sterile saline in order to avoid inflammation during the procedure and to

keep the skull cool during drilling. Any minor bleeding that occurred was remediated by

gentle application of GelFoam (Baxter, Deerfield, IL, USA) to the affected area. Once

any bleeding was stopped, GelFoam was removed and a 5mm glass coverslip was placed

over the exposed region of the brain. The coverslip was attached to the remainder of the

skull via Vetbond glue (3M, Saint Paul, MN, USA) and Ortho-Jet self-curing dental

acrylic (Lang Dental Manufacturing, Wheeling, IL, USA). Acrylic was mounded around

the outside of the coverslip to create a well for the water immersion objective that was

used for imaging through the cranial window. In addition, a stabilization bar was

mounted in the acrylic. After surgery, mice were removed from anesthesia to recover on a

warm pad before being returned to their home cage.

Two-photon confocal imaging

To visualize synaptophysin-tdtomato, mice with cranial windows were anesthetized via

isoflurane (dose) and attached to a stereotaxic instrument via the stabilization bar that

was mounted in the dental acrylic upon cranial window surgery. These mice were then

transferred to a custom built heated chamber kept at 37oC that surrounded the objective of

the microscope. Isoflurane anesthesia was maintained via dedicated lines installed in the

environmental chamber. A Leica SP5 confocal microscope fitted with a 16W Ti/Sapphire

IR laser (Chameleon, Coherent, Inc., Santa Clara, CA, USA) and a 20x Leica HCX-APO-

L, N.A. 1.0 water immersion objective was used for excitation and imaging at

wavelengths tuned between 840 and 977nm. The tip of the objective was submerged in

113

water that was pooled in the dental acrylic mound over the cranial window. The water

column was maintained by dedicated water lines included in the environmental chamber.

Time-lapse imaging was performed at a rate of 1 image per 30 seconds with 30 seconds

of rest in between images. Individual images consisted of z-projections with a step-size of

1.5m. After imaging, mice were allowed to recover in a warmed area before being

returned to their home cage.

Perfusion, sectioning, immunohistochemistry

Mice were deeply anesthetized via isoflurane inhalation and then transcardially perfused

with 1x PBS followed by 4% paraformaldahyde. Brains were then removed and drop-

fixed in 4% paraformaldahyde for 1 hour before being transferred to 30% sucrose

solution for cryoprotection. After brains were sunk in the sucrose solution, they were

frozen in dry-ice cooled 2-methylbutane and stored at -80oC until sectioning. Prior to

sectioning, tissue was mounted in OCT Compound (Tissue-Tek, The Netherlands) and

frozen on dry ice. Brains were then sectioned via Leica CM1850 cryostat at thicknesses

of either 10-15um (for antibody labeling) or 40um (for sections without antibody

labeling). Immunohistochemistry was then performed on thin (10-15um) sections

utilizing rabbit anti-synapsin primary antibodies (Synaptic Systems, Goettingen,

Germany) with anti-rabbit secondary antibodies conjugated to Alexa Fluor 488. Sections

were then mounted with fluoromount containing DABCO (Sigma, St. Louis, MO, USA)

and imaged via a C1 Plus confocal system on a Nikon Eclipse Ti-E microscope utilizing

a 20x Nikon Plan Apo 0.75NA objective.

Image analysis and statistics

114

All synaptophysin-tdtomato puncta were manually tracked via Imaris software.

Movement was analyzed via custom written programs in MATLAB. Significance was

determined via Wilcoxon rank sum test unless otherwise indicated. All error bars are

S.E.M. unless otherwise noted.

115

Chapter 5: Discussion

In Chapter 2 we hypothesized that trafficking of STVs and PTVs is coordinated

even prior to synapse assembly, which could facilitate rapid assembly. By performing

live multi-channel fluorescence confocal imaging in neurons expressing both

synaptophysin-mRFP and GFP-bassoon, we were able to record and analyze the

movements of STVs and PTVs simultaneously in the same neuron. The results indicate

that STVs and PTVs are coordinated during transport and before stabilization since PTVs

and STVs move together within the axon. We also find that STVs and PTVs pause at the

same sites within the axon, particularly when the other type of vesicle is also paused at

that site. This attraction of STVs and PTVs to the same pause sites is not mediated by a

direct interaction between STVs and PTVs since reducing the density of PTVs in the

axon yielded only small changes in STV pausing. In summary, the data support a model

of synapse formation in which STV and PTV trafficking is coordinated even prior to axo-

dendritic adhesion. This coordination includes coincident stopping of STVs and PTVs at

predefined sites of synapse formation, independent of a direct interaction between STVs

and PTVs.

Is STV and PTV transport coordinated?

Although it is clear that both PTVs and STVs need to be recruited to the same

sites in order for pre-synaptic terminals to develop, it is not immediately apparent as to

how they both arrive at the same destinations. Previously, STV and PTV dynamics have

been imaged only separately, leaving it unknown whether they are transported together

(47, 48, 50, 51, 53, 58-60, 210, 250). Recent work showed that clear, synaptic vesicle

116

protein-containing vesicles and dense-core, PTV-like vesicles can be seen apparently

tethered together in electron micrographs of young neurons (49). This was the first

indication that PTVs and STVs might be transported together, but it remained unclear

whether these vesicle aggregates correspond to either STVs and PTVs being transported

together or paused together or a later stage in synapse development. Our analysis

indicates that a sizable portion of STVs and PTVs move with each other and spend the

majority of their time together. STVs and PTVs often moved separately before or after

moving together, suggesting that STVs and PTVs can move with each other while also

maintaining their separate identities. This observation indicates that, although it is

possible that some STVs contained molecules of GFP-bassoon and/or a portion of PTVs

contained molecules of synaptophysin-mRFP, missorting of STV and PTV marker

proteins cannot account for the observed co-transport of STVs and PTVs. Coordinating

STV and PTV movement could represent a mechanism for rapid synapse development,

since a full complement of synaptic proteins could immediately be delivered to a

potential synaptic site. It will be important in the future to determine whether co-

transport is mediated by a direct interaction between STVs and PTVs or by other

mechanisms.

Previously, it was shown that STV pause sites are preferred sites of synapse

formation (51), raising the question of whether PTVs also tend to pause at these same

sites of synapse formation. By labeling both STVs and PTVs within the same axon, we

were also able to compare spatial and temporal properties of pausing for both types of

vesicles. STV and PTV pausing were qualitatively and quantitatively similar. Like

STVs, multiple PTVs paused at the same sites, and the same PTVs returned repeatedly to

117

a given site. The frequency of pausing and duration of pauses for STVs and PTVs were

also similar. Importantly, STVs and PTVs paused at the same sites within the axon. This

implies that PTVs also pause at predefined sites of synapse formation. The propensity of

STVs and PTVs to preferentially pause at the same sites within the axon suggests that

there is a mechanism that recruits both types of vesicles to the same site at the same time.

These same mechanisms could be utilized by axons to recruit the necessary proteins

during synapse formation.

What causes STV and PTV pausing?

Our data indicate that PTVs preferentially pause at sites where an STV is present.

Likewise, STVs preferentially pause at sites where a PTV is present. There are at least

three potential hypotheses which could account for this phenomenon (Figure 5.1). First,

pausing of STVs and PTVs at the same sites could be a result of STVs and PTVs

travelling together and consequently being recruited together. Second, it is possible that a

paused vesicle can interact with and stabilize a moving vesicle. In this scenario, a local

signal could cause one type of vesicle to pause at a specific site within the axon. That

paused vesicle could then interact with other vesicles and cause them to pause at the same

site. Third, STVs and PTVs could be independently attracted to a common signal within

the axon. In this case, both STVs and PTVs would respond to that signal by preferentially

pausing at the site of the signal, regardless of the presence of other vesicles. This, in turn,

would increase the probability that either vesicle was present at the site of the signal and,

therefore, the chance STVs and PTVs are simultaneously paused at the same site. Each

hypothesis represents a viable mechanism through which both synaptic vesicle and active

zone proteins could be recruited to the same site in the axon.

118

Our data are most consistent with the hypothesis that STVs and PTVs are

recruited to the same sites at the same time as a consequence of responding to a common

local signal. The first mechanism – simultaneous recruitment of STVs and PTVs that are

transported together – cannot by itself explain coincident pausing since in many cases the

STVs and PTVs that paused together paused sequentially rather than simultaneously (e.g.

see Figure 2.3A, 2.3B-1 and 2.3B-3). To distinguish between the remaining two

mechanisms, we reduced the density of PTVs within the axon, using a dominant negative

construct corresponding to the syntaxin-binding domain of syntabulin. If PTVs interact

with STVs to influence STV pausing and recruitment to a given site, then decreasing

PTV density should substantially alter STV pausing. However, in SBD-transfected axons,

STVs displayed only small differences in pause duration and no change in the frequency

of pausing when compared to STVs in control axons. Although differences in pause

duration and velocity were statistically significant, the relatively subtle change in STV

pausing suggests that PTVs exert only a small influence on STV pausing. This argues

against the second hypothesis, in which STV and PTV pausing is exclusively controlled

by an interaction between STVs and PTVs. However, based on the data, it is still possible

that STVs and PTVs respond to both local signals and one another.

It will be important in the future to identify the signals that recruit STVs and

PTVs. It is not yet known which signals cause vesicles to pause. Potential signals include

calcium, phosphorylation or small GTPase activity. Although it was previously shown

that calcium could increase the duration of pausing, the probability of pausing was

unchanged by reducing intracellular calcium (51), suggesting that calcium alone may not

be responsible. Phosphorylation is known to regulate the activity of microtubule motor

119

proteins and their association with microtubules or vesicles (101, 251-254), raising the

possibility that phosphorylation of motor proteins is the ultimate mechanism by which

local signals cause STVs and PTVs to stop their transport. Small GTPases are well-

known for their ability to control vesicle targeting (255-257), and recent work has shown

that in C. elegans, arl-8, an Arf-like small G-protein, controls where along the axon

synaptic proteins aggregate (95).

Why don’t all vesicles pause at any given site?

Although the majority of STVs and PTVs that encounter pause sites will pause at

them, many vesicles ignore the pause sites and continue their movements. This is even

true for stopping and recruitment that occurs in response to synaptogenic adhesion or

axo-dendritic contact (51). It is not clear why. One possibility is that there could be a

limited number of “docking sites” at each pause site. This seems unlikely since some

vesicles pass by even while others can still pause. Alternatively, STVs and PTVs could

exist in multiple states, at least one of which is unresponsive to local signals that cause

pausing. PTVs and STVs display heterogeneity in apparent size, velocity, and pausing

properties, which has been noted here and elsewhere (47, 48, 51, 53, 58, 60, 73). These

apparent differences may correlate with structural differences (49). In general,

smaller/dimmer vesicles appear more likely to move quickly and pass by pause sites

without pausing. Perhaps these apparently smaller vesicles are unable to respond to the

cues that induce pausing. Such a state could be a result of differences in their motors or in

other proteins associated with the STVs or PTVs. The idea of STVs and PTVs as

heterogeneous sub-populations will be an interesting topic for further exploration.

120

A model of coordinated transport

For synapses to form, all of the components of presynaptic terminals must be

delivered rapidly to the site of synapse formation. Despite its importance, we are only

beginning to understand how this occurs. Here we have proposed a model for presynaptic

terminal assembly in which trafficking of synaptic vesicle and active zone proteins is

coordinated even prior to axo-dendritic contact. This coordination occurs through a

combination of co-transport, perhaps in aggregates of vesicles tethered together (49), and

co-pausing in response to local signals. This coordination then can facilitate presynaptic

terminal assembly and contribute to the rapid recruitment of synaptic components that

has been consistently observed. The same signals that induce co-pausing prior to

transport may act down-stream of synaptogenic adhesion to simultaneously attract STVs

and PTVs to sites of synapse assembly.

121

Figure 5.1: Models for coordination of recruitment of STVs and PTVs to the same

place at the same time. STVs and PTVs could be simultaneously attracted to a given site

by (1) attraction of STVs and PTVs that are co-transported; (2) sequential recruitment via

a physical interaction between STVs and PTVs; and (3) simultaneous but independent

recruitment in response to a common signal. Our data are most consistent with the third

model.

122

What are the dynamics of presynaptic protein recruitment at developing terminals?

In Chapter 3, live, time-lapse confocal imaging of fluorescently-tagged proteins

was performed at developing presynaptic terminals which were artificially induced

through the interaction of trans-synaptic adhesion molecules. By imaging presynaptic

proteins at newly induced synapses over the course of minutes and hours, we were able to

determine that presynaptic recruitment is much more labile than previously thought. In

addition, live imaging of developing synapses in axons co-expressing synaptophysin-RFP

and neurexin-GFP established the timecourse and dynamics through which neurexin

mediates it synaptogenic signal. Furthermore, tracking STVs and PTVs outside of

synaptogenic regions strongly suggested that these components are trapped at developing

synapses, rather than actively recruited to them.

Presynaptic proteins are trapped at, but not attracted to, developing synapses

Presynaptic terminal formation is initiated by contact between axons and

dendrites, which leads to interactions between trans-synaptic adhesion molecules,

recruitment of presynaptic proteins and subsequent organization of presynaptic terminals

(Figure 5.2). It remains unclear how trans-synaptic signaling recruits presynaptic

proteins. On one hand, trans-synaptic adhesion could actively attract transport vesicles to

sites of axo-dendritic contact. Alternatively, trans-synaptic adhesion could establish sites

in the axon where synaptic proteins become trapped and organized into a presynaptic

terminal, without actively attracting transport vesicles.

Our results favor the “trapping” model (Figure 5.2, step 2, top). STVs and PTVs

did not preferentially move towards sites of trans-synaptic signaling, and their transport

123

was minimally altered by trans-synaptic adhesion. These findings suggest that trans-

synaptic adhesion does not actively attract transport vesicles into developing presynaptic

terminals. Instead, these data support the hypothesis that trans-synaptic adhesion

establishes sites in the axon that capture and stabilize available presynaptic protein.

Importantly, STVs were not attracted to SynCAM-1 or both neuroligin-1 and SynCAM-

1, and STV movement was unaffected by this combination of trans-synaptic signaling.

Therefore, assembly of synaptic proteins at developing presynaptic terminals via

establishment of sites of stabilization (rather than active attraction) appears to represent a

general mechanism that is conserved across synaptogenic pathways.

SV proteins preferentially accumulate at predefined sites that are intrinsic to the

axon (51). Although it is not known what stabilizes synaptic proteins at these specific

sites, STVs frequently pause at these sites prior to synaptogenesis. Trans-synaptic

adhesion might facilitate recruitment by stabilizing paused STVs (5, 51, 136). It is

important to note that STVs carry a variety of synaptic vesicle and some active zone

proteins (53), so it remains unclear which proteins mediate any interactions (direct or

indirect) between STVs and neurexin or SynCAM. ARL-8, which is transported with

STVs in C. elegans, has recently been shown to control aggregation and delivery of SV

and AZ proteins to appropriate sites of synaptogenesis (95, 96). Synaptogenic adhesion

could act through regulation of ARL-8 or a similar pathway to stabilize paused STVs.

The trapping model predicts that the supply of synaptic proteins in the axon

would determine the probability that synaptic proteins encounter these sites and,

therefore, levels of trapping and recruitment. Consistent with this, synapse formation is

enhanced by increasing the number of mobile STVs in the axon (160, 258), and synaptic

124

protein recruitment to poly-D-lysine coated beads preferentially occurs in axons with

high levels of protein expression and mobility (116). Interestingly, disruptions in

transport have been linked to defects in synapse formation and function (95, 96, 102, 111,

259, 260) and associated with intellectual disability in humans (261).

Synaptic proteins are labile at developing synapses

While there have been studies of the stability of new synapses (215, 262), it

remained unclear how stable synaptic proteins are at synapses after their initial

recruitment. Here, we showed that enrichment of SV proteins at individual trans-synaptic

signaling sites is highly labile: although SV proteins remained enriched at these sites, the

degree of enrichment rapidly increased and decreased, oftentimes varying over the course

of minutes. Similar fluctuations in synaptic protein levels have been observed at mature

synapses, where changes in levels of synaptic vesicles and active zone proteins have been

directly linked to changes in synapse function (263-265). At mature synapses, lability in

presynaptic protein levels arises through sharing of presynaptic components between

adjacent synapses (218-221, 223-225, 266, 267), and individual shared SVs can be

functionally integrated into the synapse within minutes after their arrival (223, 267). Our

results suggest that this dynamic arrival and departure (fluctuation) of SVs at synapses

occurs at the earliest stages in synapse development. It remains to be seen whether SVs at

developing synapses can rapidly become incorporated into the synapse and fuse with the

plasma membrane in response to action potentials. However, STVs can fuse with the

plasma membrane upon neuronal depolarization (47, 51, 58) and most likely release

glutamate at these sites when they do (51). Since glutamatergic signaling regulates the

accumulation of presynaptic proteins at developing synapses (268-270), varying the level

125

of presynaptic proteins at new excitatory synapses could alter glutamate release and

contribute to feedback or auto-regulation of synapse maturation.

We observed both gradual and sudden, dramatic changes in the amount of

synaptic proteins recruited. These two types of fluctuations in recruitment may represent

distinct features of presynaptic growth and stabilization. The sudden appearance of

synaptophysin at the site of trans-synaptic adhesion indicates that presynaptic proteins

can be recruited and stabilized at contacts on the order of seconds. This is striking

considering recent evidence that dendritic spines can form in seconds in response to

glutamate uncaging (271). Together, these data suggest that complete synapses can form

extremely rapidly. In addition, the immediate loss of synaptophysin puncta that we

observed suggests that developing synapses can be eliminated equally quickly. Finally,

the gradual accumulation or loss of synaptophysin that occurred at many contacts may

correspond to changes in presynaptic maturation or strength (272).

A model of neuroligin-neurexin mediated presynaptic protein recruitment

It has been proposed that synaptogenesis occurs in a 2-step model: postsynaptic

clusters of neuroligin induce neurexin clustering, then clustered neurexin seeds the

recruitment of presynaptic proteins (124). Our data support the hypothesis that neurexin

clustering is critical for stable synaptic protein recruitment. However, we also found that

neurexin clustering and synaptophysin recruitment appeared simultaneously. Although

the two steps of the model could occur sequentially but very rapidly, an intriguing

alternative explanation is that a cue within the axon cooperatively contributes to both

neurexin clustering and STV accumulation. Consistent with this idea, recent work in

Drosophila showed that neurexin clustering is mediated by axonal syd-1, and that this

126

process is critical for the synaptogenic ability of neurexin (125). Syd-1 acts upstream of

syd-2/liprin-α, a protein linked to transport of STVs and initiation of active zone

formation (77, 196-198, 203, 273, 274). It will be important in the future to determine

whether similar molecular mechanisms link neurexin clustering to STV stabilization in

mammals. Although mammalian orthologs of syd-1 do not contain the requisite PDZ for

neurexin organization, mammalian syd-1 (mSYD1A) has recently been shown to both

regulate SV protein accumulation during synaptogenesis and interact with liprin-α2

(206). Interestingly, in Drosophila, feedback from the postsynaptic partner also

contributes to presynaptic assembly and stabilization (125). We found that the presence

of the postsynaptic scaffolding molecule S-SCAM in neuroligin-1-expressing HEK-293

cells increased the degree of neurexin clustering in the axon, consistent with analogous

cooperative pre-and postsynaptic mechanisms existing in mammals.

While many proteins that interact with the cytoplasmic domain of neurexin have

been thought to be attractive candidates for mediating neurexin’s synaptogenic signaling

within the axon, recent work has revealed that the intracellular tail of neurexin is

completely dispensable for its synaptogenic properties (217). This suggests that neurexin

utilizes unidentified cis interactions with axonal transmembrane proteins to signal

intracellularly. It will be important to identify the cis-interacting neurexin partner(s) and

its cytosolic interactions in order to understand neuroligin/neurexin-mediated synapse

formation.

Here, we have focused on synapse assembly down-stream of neuroligin-neurexin

signaling, the prototypical synaptogenic adhesion pair. Additional factors also play a role

in regulating protein levels at presynaptic terminals, including local actin polymerization

127

(117, 234, 235, 275), signaling through BDNF/TrkB (258, 276-279) and NMDA

receptors (270), altering VGLUT1 expression (268, 269), and through other trans-

synaptic adhesion molecules not addressed in our study (136). These signals could

regulate neurexin clustering, function down-stream of neurexin clustering, or operate

through parallel mechanisms. Either way, these mechanisms likely work in concert to

dynamically regulate synapse development and function.

A model for dynamic presynaptic protein recruitment to a developing synapse

Our results strongly support a model of synapse formation where presynaptic

proteins are trapped at developing presynaptic sites, rather than actively attracted to them

(Figure 5.2, steps 1-2). Trapping tends to occur at sites of neurexin clustering, but not at

sites of diffuse neurexin recruitment, supporting the hypothesis that the spatial

arrangement of neurexin is critical for neurexin/neuroligin induced synaptogenesis.

Neurexin clustering and SV protein recruitment occur simultaneously (Figure 5.2, step 3).

Finally, at individual nascent presynaptic terminals, SV protein levels can vary within

minutes to hours, and these levels correlate with the amount of clustered neurexin. These

fluctuations might, in turn, affect later stages of synapse formation, maturation and

function.

128

Figure 5.2: Model of synaptic vesicle protein transport vesicle (STV) recruitment to

a developing presynaptic terminal. (1) Axo-dendritic contact induces diffuse neurexin

recruitment to the contact site. (2) STVs are not attracted to this site (bottom). Instead,

they are trapped as they encounter this site during normal axonal transport (top). (3) As

STVs become trapped, neurexin simultaneously clusters, resulting in the formation of a

stable site of synaptic protein accumulation.

129

In vivo synaptic dynamics

A number of non-mammalian experimental models have been developed to

address presynaptic protein recruitment and terminal formation within intact organisms

(96, 126, 280-289). In mice, two-photon confocal imaging has been used to analyze

postsynaptic spine formation and elimination within the intact brain by tracking

morphological changes of dendritic spines in GFP-filled cells (290-292). Presynaptic

bouton dynamics have also been evaluated in the mouse brain by identifying and

studying axon swellings in GFP/RFP/tdtomato-filled cells (293-296). However, the

identification of real presynaptic boutons is somewhat arbitrary as they are difficult to

distinguish from non-synaptic axonal swellings (297). Furthermore, smaller, more

dynamic synapses might be undetectable as axonal swellings.

In Chapter 4, we address these concerns though two-photon imaging of

synaptophysin-tdtomato in the brain of living mice through cranial windows. We find

that these fluorescently-tagged presynaptic proteins properly localize to synapses in the

brain, making it an excellent tool to study how presynaptic structures change over time.

Live, time-lapse imaging indicated that components such as synaptophysin-tdtomato in

many presynaptic terminals within the intact brain change over minutes via splitting,

merging, or wholesale movement. In addition, these experiments have shown that the

frequency and extent of these dynamic processes are altered over the course of

development, suggesting a possible functional role in the formation of circuits within the

intact brain.

130

A number of questions remain in regards to these dynamic presynaptic processes

which can be addressed by the type of in vivo imaging described here. For example, what

are the functional consequences of splitting, merging, or moving a presynaptic terminal

for the circuit? These experiments also only addressed the dynamics of presynaptic

terminals over approximately one week of development. Do these processes occur at

earlier timepoints? Do they occur and/or are they important in the aging brain? In

addition, the live imaging of synaptophysin-tdtomato was performed mainly in layer I of

the cortex. Are the dynamics of presynaptic terminals different in different layers?

Furthermore, are there any differences in presynaptic dynamics in different cell types

and/or cortical areas? While mice in these experiments were anesthetized, it is most likely

possible to perform similar experiments in awake behaving mice (298). It would,

therefore, be interesting to determine whether environmental stimuli affect presynaptic

dynamics. Finally, it will be critical to determine whether these dynamics are altered in

disease states.

Future outlooks – directions for future investigations

N-cadherin: trans-synaptic facilitator

An informative example of how elucidating the interactions between different

trans-synaptic adhesion systems can lead to a more complete understanding of

downstream synaptogenic signaling is illustrated by n-cadherin. N-cadherin, a

homophilic trans-synaptic adhesion protein, was one of the first trans-synaptic proteins

identified and was immediately hypothesized to be important for synapse formation. It

has since been shown that n-cadherin and its intracellular signaling through -catenin, -

131

pix, and scribble are important for localizing SVs to the presynaptic terminal through F-

actin polymerization (160, 161, 235). N-cadherin is also recruited to newly formed

synapses in zebrafish on the same time course as STVs (283). However, the importance

of n-cadherin for synaptogenesis decreases as neurons mature (161). In addition, synapse

number actually increases in -catenin knockout mice (160), possibly due to enhanced

SV distribution within the axon (258). Furthermore, n-cadherin by itself cannot induce

presynaptic terminal formation, as axonal contact with a non-neuronal cell expressing n-

cadherin does not lead to presynaptic terminal formation at the contact site as it does with

other post-synaptic adhesion proteins such as neuroligin and SynCAM (128, 157).

Despite this, n-cadherin interacts with other trans-synaptic adhesion proteins to

promote their function. For example, n-cadherin mediates the synaptogenic capability of

neuroligin in dendrites, most likely by enhancing neuroligin clustering postsynaptically

through intracellular interactions with -catenin and the synaptic scaffolding molecule

(S-SCAM) (216, 233, 299). In addition, the n-cadherin/-catenin complex interacts with

members of the leukocyte common antigen-related family receptor protein tyrosine

phosphatase (LAR-RPTP) family of proteins, which are important for the development

and organization of presynaptic terminals (300, 301). Also, n-cadherin is a substrate for

PTP, which can form a trans-synaptic adhesion complex with postsynaptically localized

TrkC to mediate synapse formation (169, 302). More work is necessary to characterize n-

cadherin’s interactions with other axonal trans-synaptic adhesion molecules. However, by

studying multiple trans-synaptic adhesion complexes in concert, we are beginning to

understand the complex and intertwined set of instructions that underlie synapse

formation.

132

Actin

Regulation of the axonal actin cytoskeleton may be an important step in

presynaptic terminal assembly down-stream of synaptogenic adhesion. Several synaptic

adhesion molecules control assembly of a localized network of F-actin at nascent

synapses, and formation of this cytoskeletal structure is required for synapse formation

(235, 236, 275, 303, 304). Actin depolymerization during early stages of development

leads to reduced synapse size and number (234), and various F-actin regulators play a

role in synapse development in mammals, Drosophila, and C. elegans (236, 305-309).

Importantly, the precise role of the actin cytoskeleton in presynaptic development is

beginning to be unraveled. It has long been known that F-actin within the presynaptic

terminal organizes synaptic vesicles and may localize them near the active zone

membrane. Prior to synapse assembly, depolymerization of actin alters the mobility of

STVs and their pausing behavior at sites of eventual synapse formation (51). F-actin can

also interact with active zone proteins and control active zone protein assembly at

synapses (275). In turn, active zone proteins, such as piccolo, can promote the formation

of presynaptic F-actin structures (193).

New technologies for studying synaptogenesis

It is clear that an important goal of developmental neuroscience is to understand

the mechanisms of synaptogenesis, but it is equally clear that many large gaps remain in

our understanding of the process of synapse formation. In many ways, the nature of

synapse formation has made this a difficult problem to attack. For example, organisms

need synapses to survive, limiting what can be discovered using traditional genetic

screens. In addition, observing the events of synapse formation can be challenging

133

because of (1) the unpredictable nature of when and where en passant synapses form in

the brain and (2) the rapid assembly of individual presynaptic terminals once the process

is initiated at any given axo-dendritic contact. Moreover, synapse elimination occurs

concurrently with synapse formation, making it impossible to determine whether

manipulations have specifically altered synapse assembly (as opposed to elimination) in

fixed tissue. Therefore, approaches such as fast, multi-channel time-lapse imaging are

necessary, in order to examine the time course of changes at individual synapses.

Application of recent advances in time-lapse imaging to the study of synapse

assembly should propel the field forward. For example, time-lapse imaging has

traditionally relied on the use of exogenously expressed proteins bearing fluorescent tags,

such as GFP, but exciting new methods have been developed that allow fluorescence

imaging of endogenous proteins (237, 310). In addition, the vast majority of studies

applying time-lapse imaging to study mammalian synaptogenesis have been performed in

culture. While it is likely that the core mechanisms of synaptogenesis are conserved in

vitro, it will be important to verify this with live in vivo imaging of presynaptic terminal

formation. Pioneering studies have examined presynaptic development in non-

mammalian models and at the mammalian neuromuscular junction (96, 126, 280-288);

however, presynaptic terminal development and synaptic protein recruitment will need to

continue to be investigated within the mammalian brain utilizing experiments like those

described in Chapter 4.

Another technical advance with promise to dramatically improve our

understanding of the mechanisms of synapse assembly is the development of high-

throughput genetic screens in mammalian neurons. For decades, elegant genetic screens

134

in Drosophila and C. elegans have provided insight into the molecular mechanisms of

synapse formation. Recently, screens have been used to identify molecules involved in

mammalian synaptogenesis, as well (169, 214, 311-313). These screens have been

particularly useful for identifying cell surface receptors that play a role in synaptogenesis.

Future screens should be helpful in identifying intracellular signals that orchestrate

synapse assembly. In addition, progress in cultivating neurons and organoids from human

induced pluripotent stem cells suggests such screens could be performed on human

neurons in the near future (314-316).

Conclusion

Synaptogenesis is a continuous process that occurs trillions of times over the

course of a human life, and its disruption can lead to devastating neurological and

psychiatric consequences. Although significant advancements into understanding

presynaptic development have been made, much more work is required. Critical to this

endeavor will be investigations into how all of these signaling mechanisms interact with

one another and lead to the dynamics observed in presynaptic development. Additionally,

insights into how disease-associated mechanisms affect the coordination of these

processes are needed. Ultimately, this will lead to a better understanding of how circuits

are established in the brain and how they are altered in disease.

135

Bibliography

1. Ramón y Cajal S. Recuerdos de mi vida. 3. ed. Madrid,: Impr. de J. Pueyo; 1923.

2. Lopez-Munoz F, Boya J, Alamo C. Neuron theory, the cornerstone of

neuroscience, on the centenary of the Nobel Prize award to Santiago Ramon y Cajal.

Brain Res Bull. 2006;70(4-6):391-405.

3. Foster M. A Textbook of Physiology. Part Three: The Central Nervious System. 7

ed. London: Macmillan and Co. Ltd.; 1897.

4. Melom JE, Littleton JT. Synapse development in health and disease. Curr Opin

Genet Dev. 2011;21(3):256-61.

5. Sudhof TC. Neuroligins and neurexins link synaptic function to cognitive disease.

Nature. 2008;455(7215):903-11. PMCID: 2673233.

6. Persico AM, Bourgeron T. Searching for ways out of the autism maze: genetic,

epigenetic and environmental clues. Trends Neurosci. 2006;29(7):349-58.

7. Mitchell KJ. The genetics of neurodevelopmental disease. Curr Opin Neurobiol.

2011;21(1):197-203.

8. Li N, Lee B, Liu RJ, Banasr M, Dwyer JM, Iwata M, et al. mTOR-dependent

synapse formation underlies the rapid antidepressant effects of NMDA antagonists.

Science. 2010;329(5994):959-64. PMCID: 3116441.

9. Bambico FR, Belzung C. Novel insights into depression and antidepressants: a

synergy between synaptogenesis and neurogenesis? Curr Top Behav Neurosci.

2013;15:243-91.

10. Szumlinski KK, Kalivas PW, Worley PF. Homer proteins: implications for

neuropsychiatric disorders. Curr Opin Neurobiol. 2006;16(3):251-7.

136

11. Mandela P, Ma XM. Kalirin, a key player in synapse formation, is implicated in

human diseases. Neural Plast. 2012;2012:728161. PMCID: 3324156.

12. Milnerwood AJ, Raymond LA. Early synaptic pathophysiology in

neurodegeneration: insights from Huntington's disease. Trends Neurosci.

2010;33(11):513-23.

13. Picconi B, Piccoli G, Calabresi P. Synaptic dysfunction in Parkinson's disease.

Adv Exp Med Biol. 2012;970:553-72.

14. Terry RD, Masliah E, Salmon DP, Butters N, DeTeresa R, Hill R, et al. Physical

basis of cognitive alterations in Alzheimer's disease: synapse loss is the major correlate of

cognitive impairment. Ann Neurol. 1991;30(4):572-80.

15. Clare R, King VG, Wirenfeldt M, Vinters HV. Synapse loss in dementias. J

Neurosci Res. 2010;88(10):2083-90. PMCID: 3068914.

16. Lu B, Nagappan G, Guan X, Nathan PJ, Wren P. BDNF-based synaptic repair as

a disease-modifying strategy for neurodegenerative diseases. Nat Rev Neurosci.

2013;14(6):401-16.

17. Connors BW, Long MA. Electrical synapses in the mammalian brain. Annu Rev

Neurosci. 2004;27:393-418.

18. Goodenough DA, Paul DL. Gap junctions. Cold Spring Harb Perspect Biol.

2009;1(1):a002576. PMCID: 2742079.

19. Saez JC, Berthoud VM, Branes MC, Martinez AD, Beyer EC. Plasma membrane

channels formed by connexins: their regulation and functions. Physiol Rev.

2003;83(4):1359-400.

137

20. Laird DW. Life cycle of connexins in health and disease. Biochem J. 2006;394(Pt

3):527-43. PMCID: 1383703.

21. Bennett MV, Zukin RS. Electrical coupling and neuronal synchronization in the

Mammalian brain. Neuron. 2004;41(4):495-511.

22. Sudhof TC, Jahn R. Proteins of synaptic vesicles involved in exocytosis and

membrane recycling. Neuron. 1991;6(5):665-77.

23. Jahn R, Sudhof TC. Synaptic vesicles and exocytosis. Annu Rev Neurosci.

1994;17:219-46.

24. Sudhof TC. The synaptic vesicle cycle. Annu Rev Neurosci. 2004;27:509-47.

25. Sudhof TC, Rizo J. Synaptic vesicle exocytosis. Cold Spring Harb Perspect Biol.

2011;3(12). PMCID: 3225952.

26. Barnes NM, Sharp T. A review of central 5-HT receptors and their function.

Neuropharmacology. 1999;38(8):1083-152.

27. Burnstock G. Pathophysiology and therapeutic potential of purinergic signaling.

Pharmacol Rev. 2006;58(1):58-86.

28. Carlsson A. Perspectives on the discovery of central monoaminergic

neurotransmission. Annu Rev Neurosci. 1987;10:19-40.

29. Lerma J. Kainate receptor physiology. Curr Opin Pharmacol. 2006;6(1):89-97.

30. Perry E, Walker M, Grace J, Perry R. Acetylcholine in mind: a neurotransmitter

correlate of consciousness? Trends Neurosci. 1999;22(6):273-80.

31. Wilson RI, Nicoll RA. Endocannabinoid signaling in the brain. Science.

2002;296(5568):678-82.

138

32. Schwartz JC, Arrang JM, Garbarg M, Pollard H, Ruat M. Histaminergic

transmission in the mammalian brain. Physiol Rev. 1991;71(1):1-51.

33. Emson PC, Lindvall O. Distribution of putative neurotransmitters in the

neocortex. Neuroscience. 1979;4(1):1-30.

34. Pakkenberg B, Gundersen HJ. Neocortical neuron number in humans: effect of

sex and age. J Comp Neurol. 1997;384(2):312-20.

35. Pakkenberg B, Pelvig D, Marner L, Bundgaard MJ, Gundersen HJ, Nyengaard

JR, et al. Aging and the human neocortex. Exp Gerontol. 2003;38(1-2):95-9.

36. Azevedo FA, Carvalho LR, Grinberg LT, Farfel JM, Ferretti RE, Leite RE, et al.

Equal numbers of neuronal and nonneuronal cells make the human brain an isometrically

scaled-up primate brain. J Comp Neurol. 2009;513(5):532-41.

37. Wilhelm BG, Mandad S, Truckenbrodt S, Krohnert K, Schafer C, Rammner B, et

al. Composition of isolated synaptic boutons reveals the amounts of vesicle trafficking

proteins. Science. 2014;344(6187):1023-8.

38. Siksou L, Triller A, Marty S. Ultrastructural organization of presynaptic

terminals. Curr Opin Neurobiol. 2011;21(2):261-8.

39. Harris KM, Weinberg RJ. Ultrastructure of synapses in the mammalian brain.

Cold Spring Harb Perspect Biol. 2012;4(5). PMCID: 3331701.

40. Harris KM, Sultan P. Variation in the number, location and size of synaptic

vesicles provides an anatomical basis for the nonuniform probability of release at

hippocampal CA1 synapses. Neuropharmacology. 1995;34(11):1387-95.

41. Schikorski T, Stevens CF. Quantitative ultrastructural analysis of hippocampal

excitatory synapses. J Neurosci. 1997;17(15):5858-67.

139

42. Qu L, Akbergenova Y, Hu Y, Schikorski T. Synapse-to-synapse variation in mean

synaptic vesicle size and its relationship with synaptic morphology and function. J Comp

Neurol. 2009;514(4):343-52.

43. Zhang B, Koh YH, Beckstead RB, Budnik V, Ganetzky B, Bellen HJ. Synaptic

vesicle size and number are regulated by a clathrin adaptor protein required for

endocytosis. Neuron. 1998;21(6):1465-75.

44. Sudhof TC. The presynaptic active zone. Neuron. 2012;75(1):11-25. PMCID:

3743085.

45. McAllister AK. Dynamic aspects of CNS synapse formation. Annu Rev Neurosci.

2007;30:425-50. PMCID: 3251656.

46. Bury LA, Sabo SL. How it's made: the synapse. Mol Interv. 2010;10(5):282-92.

47. Kraszewski K, Mundigl O, Daniell L, Verderio C, Matteoli M, De Camilli P.

Synaptic vesicle dynamics in living cultured hippocampal neurons visualized with CY3-

conjugated antibodies directed against the lumenal domain of synaptotagmin. J Neurosci.

1995;15(6):4328-42.

48. Nakata T, Terada S, Hirokawa N. Visualization of the dynamics of synaptic

vesicle and plasma membrane proteins in living axons. J Cell Biol. 1998;140(3):659-74.

PMCID: 2140163.

49. Tao-Cheng JH. Ultrastructural localization of active zone and synaptic vesicle

proteins in a preassembled multi-vesicle transport aggregate. Neuroscience.

2007;150(3):575-84. PMCID: 2190624.

50. Sabo SL, McAllister AK. Mobility and cycling of synaptic protein-containing

vesicles in axonal growth cone filopodia. Nat Neurosci. 2003;6(12):1264-9.

140

51. Sabo SL, Gomes RA, McAllister AK. Formation of presynaptic terminals at

predefined sites along axons. J Neurosci. 2006;26(42):10813-25.

52. Okada Y, Yamazaki H, Sekine-Aizawa Y, Hirokawa N. The neuron-specific

kinesin superfamily protein KIF1A is a unique monomeric motor for anterograde axonal

transport of synaptic vesicle precursors. Cell. 1995;81(5):769-80.

53. Ahmari SE, Buchanan J, Smith SJ. Assembly of presynaptic active zones from

cytoplasmic transport packets. Nat Neurosci. 2000;3(5):445-51.

54. Verderio C, Coco S, Bacci A, Rossetto O, De Camilli P, Montecucco C, et al.

Tetanus toxin blocks the exocytosis of synaptic vesicles clustered at synapses but not of

synaptic vesicles in isolated axons. J Neurosci. 1999;19(16):6723-32.

55. Zakharenko S, Chang S, O'Donoghue M, Popov SV. Neurotransmitter secretion

along growing nerve processes: comparison with synaptic vesicle exocytosis. J Cell Biol.

1999;144(3):507-18. PMCID: 2132923.

56. Krueger SR, Kolar A, Fitzsimonds RM. The presynaptic release apparatus is

functional in the absence of dendritic contact and highly mobile within isolated axons.

Neuron. 2003;40(5):945-57.

57. Matteoli M, Takei K, Perin MS, Sudhof TC, De Camilli P. Exo-endocytotic

recycling of synaptic vesicles in developing processes of cultured hippocampal neurons. J

Cell Biol. 1992;117(4):849-61. PMCID: 2289460.

58. Dai Z, Peng HB. Dynamics of synaptic vesicles in cultured spinal cord neurons in

relationship to synaptogenesis. Mol Cell Neurosci. 1996;7(6):443-52.

141

59. Zhai RG, Vardinon-Friedman H, Cases-Langhoff C, Becker B, Gundelfinger ED,

Ziv NE, et al. Assembling the presynaptic active zone: a characterization of an active one

precursor vesicle. Neuron. 2001;29(1):131-43.

60. Shapira M, Zhai RG, Dresbach T, Bresler T, Torres VI, Gundelfinger ED, et al.

Unitary assembly of presynaptic active zones from Piccolo-Bassoon transport vesicles.

Neuron. 2003;38(2):237-52.

61. Dresbach T, Torres V, Wittenmayer N, Altrock WD, Zamorano P, Zuschratter W,

et al. Assembly of active zone precursor vesicles: obligatory trafficking of presynaptic

cytomatrix proteins Bassoon and Piccolo via a trans-Golgi compartment. J Biol Chem.

2006;281(9):6038-47.

62. Maas C, Torres VI, Altrock WD, Leal-Ortiz S, Wagh D, Terry-Lorenzo RT, et al.

Formation of Golgi-derived active zone precursor vesicles. J Neurosci.

2012;32(32):11095-108.

63. Goldstein LS, Yang Z. Microtubule-based transport systems in neurons: the roles

of kinesins and dyneins. Annu Rev Neurosci. 2000;23:39-71.

64. Heidemann SR, Landers JM, Hamborg MA. Polarity orientation of axonal

microtubules. J Cell Biol. 1981;91(3 Pt 1):661-5. PMCID: 2112798.

65. Burton PR, Paige JL. Polarity of axoplasmic microtubules in the olfactory nerve

of the frog. Proc Natl Acad Sci U S A. 1981;78(5):3269-73. PMCID: 319543.

66. Baas PW, Deitch JS, Black MM, Banker GA. Polarity orientation of microtubules

in hippocampal neurons: uniformity in the axon and nonuniformity in the dendrite. Proc

Natl Acad Sci U S A. 1988;85(21):8335-9. PMCID: 282424.

142

67. Baas PW, Lin S. Hooks and comets: The story of microtubule polarity orientation

in the neuron. Dev Neurobiol. 2011;71(6):403-18. PMCID: 3151545.

68. Yonekawa Y, Harada A, Okada Y, Funakoshi T, Kanai Y, Takei Y, et al. Defect

in synaptic vesicle precursor transport and neuronal cell death in KIF1A motor protein-

deficient mice. J Cell Biol. 1998;141(2):431-41. PMCID: 2148442.

69. Hall DH, Hedgecock EM. Kinesin-related gene unc-104 is required for axonal

transport of synaptic vesicles in C. elegans. Cell. 1991;65(5):837-47.

70. Nakamura N, Miyake Y, Matsushita M, Tanaka S, Inoue H, Kanazawa H.

KIF1Bbeta2, capable of interacting with CHP, is localized to synaptic vesicles. J

Biochem. 2002;132(3):483-91.

71. Zhao C, Takita J, Tanaka Y, Setou M, Nakagawa T, Takeda S, et al. Charcot-

Marie-Tooth disease type 2A caused by mutation in a microtubule motor KIF1Bbeta.

Cell. 2001;105(5):587-97.

72. Sgro AE, Bajjalieh SM, Chiu DT. Single-axonal organelle analysis method

reveals new protein-motor associations. ACS Chem Neurosci. 2013;4(2):277-84.

PMCID: 3751541.

73. Cai Q, Pan PY, Sheng ZH. Syntabulin-kinesin-1 family member 5B-mediated

axonal transport contributes to activity-dependent presynaptic assembly. J Neurosci.

2007;27(27):7284-96.

74. Su Q, Cai Q, Gerwin C, Smith CL, Sheng ZH. Syntabulin is a microtubule-

associated protein implicated in syntaxin transport in neurons. Nat Cell Biol.

2004;6(10):941-53.

143

75. Sato-Yoshitake R, Yorifuji H, Inagaki M, Hirokawa N. The phosphorylation of

kinesin regulates its binding to synaptic vesicles. J Biol Chem. 1992;267(33):23930-6.

76. Hurd DD, Saxton WM. Kinesin mutations cause motor neuron disease phenotypes

by disrupting fast axonal transport in Drosophila. Genetics. 1996;144(3):1075-85.

PMCID: 1207603.

77. Miller KE, DeProto J, Kaufmann N, Patel BN, Duckworth A, Van Vactor D.

Direct observation demonstrates that Liprin-alpha is required for trafficking of synaptic

vesicles. Curr Biol. 2005;15(7):684-9.

78. Byrd DT, Kawasaki M, Walcoff M, Hisamoto N, Matsumoto K, Jin Y. UNC-16, a

JNK-signaling scaffold protein, regulates vesicle transport in C. elegans. Neuron.

2001;32(5):787-800.

79. Takamori S, Holt M, Stenius K, Lemke EA, Gronborg M, Riedel D, et al.

Molecular anatomy of a trafficking organelle. Cell. 2006;127(4):831-46.

80. Vallee RB, Shpetner HS, Paschal BM. The role of dynein in retrograde axonal

transport. Trends Neurosci. 1989;12(2):66-70.

81. Waterman-Storer CM, Karki SB, Kuznetsov SA, Tabb JS, Weiss DG, Langford

GM, et al. The interaction between cytoplasmic dynein and dynactin is required for fast

axonal transport. Proc Natl Acad Sci U S A. 1997;94(22):12180-5. PMCID: 23743.

82. Hirokawa N, Sato-Yoshitake R, Yoshida T, Kawashima T. Brain dynein

(MAP1C) localizes on both anterogradely and retrogradely transported membranous

organelles in vivo. J Cell Biol. 1990;111(3):1027-37. PMCID: 2116262.

144

83. Fejtova A, Davydova D, Bischof F, Lazarevic V, Altrock WD, Romorini S, et al.

Dynein light chain regulates axonal trafficking and synaptic levels of Bassoon. J Cell

Biol. 2009;185(2):341-55. PMCID: 2700376.

84. Vallee RB, Williams JC, Varma D, Barnhart LE. Dynein: An ancient motor

protein involved in multiple modes of transport. J Neurobiol. 2004;58(2):189-200.

85. Koushika SP, Schaefer AM, Vincent R, Willis JH, Bowerman B, Nonet ML.

Mutations in Caenorhabditis elegans cytoplasmic dynein components reveal specificity of

neuronal retrograde cargo. J Neurosci. 2004;24(16):3907-16.

86. Li JY, Pfister KK, Brady ST, Dahlstrom A. Cytoplasmic dynein conversion at a

crush injury in rat peripheral axons. J Neurosci Res. 2000;61(2):151-61.

87. Bury LA, Sabo SL. Coordinated trafficking of synaptic vesicle and active zone

proteins prior to synapse formation. Neural Dev. 2011;6:24. PMCID: 3103415.

88. Bury LA, Sabo SL. Dynamic mechanisms of neuroligin-dependent presynaptic

terminal assembly in living cortical neurons. Neural Dev. 2014;9:13. PMCID: 4049477.

89. Wagner OI, Esposito A, Kohler B, Chen CW, Shen CP, Wu GH, et al. Synaptic

scaffolding protein SYD-2 clusters and activates kinesin-3 UNC-104 in C. elegans. Proc

Natl Acad Sci U S A. 2009;106(46):19605-10. PMCID: 2780759.

90. Shin H, Wyszynski M, Huh KH, Valtschanoff JG, Lee JR, Ko J, et al. Association

of the kinesin motor KIF1A with the multimodular protein liprin-alpha. J Biol Chem.

2003;278(13):11393-401.

91. Niwa S, Tanaka Y, Hirokawa N. KIF1Bbeta- and KIF1A-mediated axonal

transport of presynaptic regulator Rab3 occurs in a GTP-dependent manner through

DENN/MADD. Nat Cell Biol. 2008;10(11):1269-79.

145

92. Wada M, Nakanishi H, Satoh A, Hirano H, Obaishi H, Matsuura Y, et al.

Isolation and characterization of a GDP/GTP exchange protein specific for the Rab3

subfamily small G proteins. J Biol Chem. 1997;272(7):3875-8.

93. Tanaka M, Miyoshi J, Ishizaki H, Togawa A, Ohnishi K, Endo K, et al. Role of

Rab3 GDP/GTP exchange protein in synaptic vesicle trafficking at the mouse

neuromuscular junction. Mol Biol Cell. 2001;12(5):1421-30. PMCID: 34594.

94. Yamaguchi K, Tanaka M, Mizoguchi A, Hirata Y, Ishizaki H, Kaneko K, et al. A

GDP/GTP exchange protein for the Rab3 small G protein family up-regulates a

postdocking step of synaptic exocytosis in central synapses. Proc Natl Acad Sci U S A.

2002;99(22):14536-41. PMCID: 137918.

95. Klassen MP, Wu YE, Maeder CI, Nakae I, Cueva JG, Lehrman EK, et al. An Arf-

like small G protein, ARL-8, promotes the axonal transport of presynaptic cargoes by

suppressing vesicle aggregation. Neuron. 2010;66(5):710-23. PMCID: 3168544.

96. Wu YE, Huo L, Maeder CI, Feng W, Shen K. The balance between capture and

dissociation of presynaptic proteins controls the spatial distribution of synapses. Neuron.

2013;78(6):994-1011.

97. Zheng Q, Ahlawat S, Schaefer A, Mahoney T, Koushika SP, Nonet ML. The

vesicle protein SAM-4 regulates the processivity of synaptic vesicle transport. PLoS

Genet. 2014;10(10):e1004644. PMCID: 4199485.

98. Klopfenstein DR, Tomishige M, Stuurman N, Vale RD. Role of

phosphatidylinositol(4,5)bisphosphate organization in membrane transport by the Unc104

kinesin motor. Cell. 2002;109(3):347-58. PMCID: 2851634.

146

99. Klopfenstein DR, Vale RD. The lipid binding pleckstrin homology domain in

UNC-104 kinesin is necessary for synaptic vesicle transport in Caenorhabditis elegans.

Mol Biol Cell. 2004;15(8):3729-39. PMCID: 491832.

100. Kardon JR, Vale RD. Regulators of the cytoplasmic dynein motor. Nat Rev Mol

Cell Biol. 2009;10(12):854-65. PMCID: 3394690.

101. Okada Y, Sato-Yoshitake R, Hirokawa N. The activation of protein kinase A

pathway selectively inhibits anterograde axonal transport of vesicles but not mitochondria

transport or retrograde transport in vivo. J Neurosci. 1995;15(4):3053-64.

102. Verhey KJ, Hammond JW. Traffic control: regulation of kinesin motors. Nat Rev

Mol Cell Biol. 2009;10(11):765-77.

103. Schlager MA, Hoogenraad CC. Basic mechanisms for recognition and transport

of synaptic cargos. Mol Brain. 2009;2:25. PMCID: 2732917.

104. Fukushima N, Furuta D, Hidaka Y, Moriyama R, Tsujiuchi T. Post-translational

modifications of tubulin in the nervous system. J Neurochem. 2009;109(3):683-93.

105. Garnham CP, Roll-Mecak A. The chemical complexity of cellular microtubules:

tubulin post-translational modification enzymes and their roles in tuning microtubule

functions. Cytoskeleton (Hoboken). 2012;69(7):442-63. PMCID: 3459347.

106. Baas PW, Qiang L. Neuronal microtubules: when the MAP is the roadblock.

Trends Cell Biol. 2005;15(4):183-7.

107. Duncan JE, Goldstein LS. The genetics of axonal transport and axonal transport

disorders. PLoS Genet. 2006;2(9):e124. PMCID: 1584265.

147

108. Hiruma H, Katakura T, Takahashi S, Ichikawa T, Kawakami T. Glutamate and

amyloid beta-protein rapidly inhibit fast axonal transport in cultured rat hippocampal

neurons by different mechanisms. J Neurosci. 2003;23(26):8967-77.

109. Hancock WO. Bidirectional cargo transport: moving beyond tug of war. Nat Rev

Mol Cell Biol. 2014;15(9):615-28.

110. Arnold DB. Actin and microtubule-based cytoskeletal cues direct polarized

targeting of proteins in neurons. Sci Signal. 2009;2(83):pe49. PMCID: 2937173.

111. Hirokawa N, Niwa S, Tanaka Y. Molecular motors in neurons: transport

mechanisms and roles in brain function, development, and disease. Neuron.

2010;68(4):610-38.

112. Nelson JC, Stavoe AK, Colon-Ramos DA. The actin cytoskeleton in presynaptic

assembly. Cell Adh Migr. 2013;7(4):379-87. PMCID: 3739815.

113. Bridgman PC. Myosin-dependent transport in neurons. J Neurobiol.

2004;58(2):164-74.

114. Kondo M, Takei Y, Hirokawa N. Motor protein KIF1A is essential for

hippocampal synaptogenesis and learning enhancement in an enriched environment.

Neuron. 2012;73(4):743-57.

115. Lee H, Dean C, Isacoff E. Alternative splicing of neuroligin regulates the rate of

presynaptic differentiation. J Neurosci. 2010;30(34):11435-46.

116. Suarez F, Thostrup P, Colman D, Grutter P. Dynamics of presynaptic protein

recruitment induced by local presentation of artificial adhesive contacts. Dev Neurobiol.

2013;73(1):98-106. PMCID: 3518747.

148

117. Lucido AL, Suarez Sanchez F, Thostrup P, Kwiatkowski AV, Leal-Ortiz S,

Gopalakrishnan G, et al. Rapid assembly of functional presynaptic boutons triggered by

adhesive contacts. J Neurosci. 2009;29(40):12449-66.

118. Friedman HV, Bresler T, Garner CC, Ziv NE. Assembly of new individual

excitatory synapses: time course and temporal order of synaptic molecule recruitment.

Neuron. 2000;27(1):57-69.

119. Okabe S, Miwa A, Okado H. Spine formation and correlated assembly of

presynaptic and postsynaptic molecules. J Neurosci. 2001;21(16):6105-14.

120. Ratnayaka A, Marra V, Branco T, Staras K. Extrasynaptic vesicle recycling in

mature hippocampal neurons. Nat Commun. 2011;2:531. PMCID: 3492933.

121. Knott GW, Holtmaat A, Wilbrecht L, Welker E, Svoboda K. Spine growth

precedes synapse formation in the adult neocortex in vivo. Nat Neurosci. 2006;9(9):1117-

24.

122. Nagerl UV, Kostinger G, Anderson JC, Martin KA, Bonhoeffer T. Protracted

synaptogenesis after activity-dependent spinogenesis in hippocampal neurons. J

Neurosci. 2007;27(30):8149-56.

123. Ziv NE, Garner CC. Cellular and molecular mechanisms of presynaptic assembly.

Nat Rev Neurosci. 2004;5(5):385-99.

124. Dean C, Scholl FG, Choih J, DeMaria S, Berger J, Isacoff E, et al. Neurexin

mediates the assembly of presynaptic terminals. Nat Neurosci. 2003;6(7):708-16.

PMCID: 1646425.

149

125. Owald D, Khorramshahi O, Gupta VK, Banovic D, Depner H, Fouquet W, et al.

Cooperation of Syd-1 with Neurexin synchronizes pre- with postsynaptic assembly. Nat

Neurosci. 2012;15(9):1219-26.

126. Easley-Neal C, Fierro J, Jr., Buchanan J, Washbourne P. Late recruitment of

synapsin to nascent synapses is regulated by Cdk5. Cell Rep. 2013;3(4):1199-212.

PMCID: 3742072.

127. Yeow MB, Peterson EH. Active zone organization and vesicle content scale with

bouton size at a vertebrate central synapse. J Comp Neurol. 1991;307(3):475-86.

128. Scheiffele P, Fan J, Choih J, Fetter R, Serafini T. Neuroligin expressed in

nonneuronal cells triggers presynaptic development in contacting axons. Cell.

2000;101(6):657-69.

129. Biederer T, Sara Y, Mozhayeva M, Atasoy D, Liu X, Kavalali ET, et al.

SynCAM, a synaptic adhesion molecule that drives synapse assembly. Science.

2002;297(5586):1525-31.

130. Ziv NE, Smith SJ. Evidence for a role of dendritic filopodia in synaptogenesis and

spine formation. Neuron. 1996;17(1):91-102.

131. Konur S, Yuste R. Imaging the motility of dendritic protrusions and axon

terminals: roles in axon sampling and synaptic competition. Mol Cell Neurosci.

2004;27(4):427-40.

132. Arstikaitis P, Gauthier-Campbell C, Huang K, El-Husseini A, Murphy TH.

Proteins that promote filopodia stability, but not number, lead to more axonal-dendritic

contacts. PLoS One. 2011;6(3):e16998. PMCID: 3049770.

150

133. Lohmann C, Bonhoeffer T. A role for local calcium signaling in rapid synaptic

partner selection by dendritic filopodia. Neuron. 2008;59(2):253-60.

134. De Roo M, Klauser P, Mendez P, Poglia L, Muller D. Activity-dependent PSD

formation and stabilization of newly formed spines in hippocampal slice cultures. Cereb

Cortex. 2008;18(1):151-61.

135. De Roo M, Klauser P, Muller D. LTP promotes a selective long-term stabilization

and clustering of dendritic spines. PLoS Biol. 2008;6(9):e219. PMCID: 2531136.

136. Missler M, Sudhof TC, Biederer T. Synaptic cell adhesion. Cold Spring Harb

Perspect Biol. 2012;4(4):a005694. PMCID: 3312681.

137. Akins MR, Biederer T. Cell-cell interactions in synaptogenesis. Curr Opin

Neurobiol. 2006;16(1):83-9.

138. Tallafuss A, Constable JR, Washbourne P. Organization of central synapses by

adhesion molecules. Eur J Neurosci. 2010;32(2):198-206.

139. Song JY, Ichtchenko K, Sudhof TC, Brose N. Neuroligin 1 is a postsynaptic cell-

adhesion molecule of excitatory synapses. Proc Natl Acad Sci U S A. 1999;96(3):1100-5.

PMCID: 15357.

140. Varoqueaux F, Jamain S, Brose N. Neuroligin 2 is exclusively localized to

inhibitory synapses. Eur J Cell Biol. 2004;83(9):449-56.

141. Graf ER, Kang Y, Hauner AM, Craig AM. Structure function and splice site

analysis of the synaptogenic activity of the neurexin-1 beta LNS domain. J Neurosci.

2006;26(16):4256-65. PMCID: 2826202.

151

142. Futai K, Doty CD, Baek B, Ryu J, Sheng M. Specific trans-synaptic interaction

with inhibitory interneuronal neurexin underlies differential ability of neuroligins to

induce functional inhibitory synapses. J Neurosci. 2013;33(8):3612-23.

143. Fogel AI, Akins MR, Krupp AJ, Stagi M, Stein V, Biederer T. SynCAMs

organize synapses through heterophilic adhesion. J Neurosci. 2007;27(46):12516-30.

144. Thomas LA, Akins MR, Biederer T. Expression and adhesion profiles of

SynCAM molecules indicate distinct neuronal functions. J Comp Neurol.

2008;510(1):47-67. PMCID: 2543944.

145. Ichtchenko K, Nguyen T, Sudhof TC. Structures, alternative splicing, and

neurexin binding of multiple neuroligins. J Biol Chem. 1996;271(5):2676-82.

146. Schreiner D, Nguyen TM, Russo G, Heber S, Patrignani A, Ahrne E, et al.

Targeted combinatorial alternative splicing generates brain region-specific repertoires of

neurexins. Neuron. 2014;84(2):386-98.

147. Treutlein B, Gokce O, Quake SR, Sudhof TC. Cartography of neurexin alternative

splicing mapped by single-molecule long-read mRNA sequencing. Proc Natl Acad Sci U

S A. 2014;111(13):E1291-9. PMCID: 3977267.

148. Ullrich B, Ushkaryov YA, Sudhof TC. Cartography of neurexins: more than 1000

isoforms generated by alternative splicing and expressed in distinct subsets of neurons.

Neuron. 1995;14(3):497-507.

149. Chih B, Gollan L, Scheiffele P. Alternative splicing controls selective trans-

synaptic interactions of the neuroligin-neurexin complex. Neuron. 2006;51(2):171-8.

152

150. Boucard AA, Chubykin AA, Comoletti D, Taylor P, Sudhof TC. A splice code for

trans-synaptic cell adhesion mediated by binding of neuroligin 1 to alpha- and beta-

neurexins. Neuron. 2005;48(2):229-36.

151. Traunmuller L, Bornmann C, Scheiffele P. Alternative splicing coupled nonsense-

mediated decay generates neuronal cell type-specific expression of SLM proteins. J

Neurosci. 2014;34(50):16755-61.

152. Iijima T, Iijima Y, Witte H, Scheiffele P. Neuronal cell type-specific alternative

splicing is regulated by the KH domain protein SLM1. J Cell Biol. 2014;204(3):331-42.

PMCID: 3912522.

153. Varoqueaux F, Aramuni G, Rawson RL, Mohrmann R, Missler M, Gottmann K,

et al. Neuroligins determine synapse maturation and function. Neuron. 2006;51(6):741-

54.

154. Missler M, Zhang W, Rohlmann A, Kattenstroth G, Hammer RE, Gottmann K, et

al. Alpha-neurexins couple Ca2+ channels to synaptic vesicle exocytosis. Nature.

2003;423(6943):939-48.

155. Kwon HB, Kozorovitskiy Y, Oh WJ, Peixoto RT, Akhtar N, Saulnier JL, et al.

Neuroligin-1-dependent competition regulates cortical synaptogenesis and synapse

number. Nat Neurosci. 2012;15(12):1667-74. PMCID: 3536444.

156. Robbins EM, Krupp AJ, Perez de Arce K, Ghosh AK, Fogel AI, Boucard A, et al.

SynCAM 1 adhesion dynamically regulates synapse number and impacts plasticity and

learning. Neuron. 2010;68(5):894-906. PMCID: 3026433.

153

157. Sara Y, Biederer T, Atasoy D, Chubykin A, Mozhayeva MG, Sudhof TC, et al.

Selective capability of SynCAM and neuroligin for functional synapse assembly. J

Neurosci. 2005;25(1):260-70.

158. Murase S, Mosser E, Schuman EM. Depolarization drives beta-Catenin into

neuronal spines promoting changes in synaptic structure and function. Neuron.

2002;35(1):91-105.

159. Togashi H, Abe K, Mizoguchi A, Takaoka K, Chisaka O, Takeichi M. Cadherin

regulates dendritic spine morphogenesis. Neuron. 2002;35(1):77-89.

160. Bamji SX, Shimazu K, Kimes N, Huelsken J, Birchmeier W, Lu B, et al. Role of

beta-catenin in synaptic vesicle localization and presynaptic assembly. Neuron.

2003;40(4):719-31. PMCID: 2757419.

161. Bozdagi O, Valcin M, Poskanzer K, Tanaka H, Benson DL. Temporally distinct

demands for classic cadherins in synapse formation and maturation. Mol Cell Neurosci.

2004;27(4):509-21. PMCID: 2910522.

162. Wang Z, Wang B, Yang L, Guo Q, Aithmitti N, Songyang Z, et al. Presynaptic

and postsynaptic interaction of the amyloid precursor protein promotes peripheral and

central synaptogenesis. J Neurosci. 2009;29(35):10788-801. PMCID: 2757256.

163. Ichtchenko K, Hata Y, Nguyen T, Ullrich B, Missler M, Moomaw C, et al.

Neuroligin 1: a splice site-specific ligand for beta-neurexins. Cell. 1995;81(3):435-43.

164. Nguyen T, Sudhof TC. Binding properties of neuroligin 1 and neurexin 1beta

reveal function as heterophilic cell adhesion molecules. J Biol Chem.

1997;272(41):26032-9.

154

165. de Wit J, Sylwestrak E, O'Sullivan ML, Otto S, Tiglio K, Savas JN, et al.

LRRTM2 interacts with Neurexin1 and regulates excitatory synapse formation. Neuron.

2009;64(6):799-806. PMCID: 2829666.

166. Ko J, Fuccillo MV, Malenka RC, Sudhof TC. LRRTM2 functions as a neurexin

ligand in promoting excitatory synapse formation. Neuron. 2009;64(6):791-8. PMCID:

2829314.

167. Uemura T, Lee SJ, Yasumura M, Takeuchi T, Yoshida T, Ra M, et al. Trans-

synaptic interaction of GluRdelta2 and Neurexin through Cbln1 mediates synapse

formation in the cerebellum. Cell. 2010;141(6):1068-79.

168. Lim SH, Kwon SK, Lee MK, Moon J, Jeong DG, Park E, et al. Synapse formation

regulated by protein tyrosine phosphatase receptor T through interaction with cell

adhesion molecules and Fyn. EMBO J. 2009;28(22):3564-78. PMCID: 2782100.

169. Takahashi H, Arstikaitis P, Prasad T, Bartlett TE, Wang YT, Murphy TH, et al.

Postsynaptic TrkC and presynaptic PTPsigma function as a bidirectional excitatory

synaptic organizing complex. Neuron. 2011;69(2):287-303. PMCID: 3056349.

170. Valnegri P, Montrasio C, Brambilla D, Ko J, Passafaro M, Sala C. The X-linked

intellectual disability protein IL1RAPL1 regulates excitatory synapse formation by

binding PTPdelta and RhoGAP2. Hum Mol Genet. 2011;20(24):4797-809. PMCID:

3221541.

171. Yoshida T, Yasumura M, Uemura T, Lee SJ, Ra M, Taguchi R, et al. IL-1

receptor accessory protein-like 1 associated with mental retardation and autism mediates

synapse formation by trans-synaptic interaction with protein tyrosine phosphatase delta. J

Neurosci. 2011;31(38):13485-99.

155

172. Yoshida T, Shiroshima T, Lee SJ, Yasumura M, Uemura T, Chen X, et al.

Interleukin-1 receptor accessory protein organizes neuronal synaptogenesis as a cell

adhesion molecule. J Neurosci. 2012;32(8):2588-600.

173. Dalva MB, Takasu MA, Lin MZ, Shamah SM, Hu L, Gale NW, et al. EphB

receptors interact with NMDA receptors and regulate excitatory synapse formation. Cell.

2000;103(6):945-56.

174. Kayser MS, McClelland AC, Hughes EG, Dalva MB. Intracellular and trans-

synaptic regulation of glutamatergic synaptogenesis by EphB receptors. J Neurosci.

2006;26(47):12152-64.

175. Silva JP, Lelianova VG, Ermolyuk YS, Vysokov N, Hitchen PG, Berninghausen

O, et al. Latrophilin 1 and its endogenous ligand Lasso/teneurin-2 form a high-affinity

transsynaptic receptor pair with signaling capabilities. Proc Natl Acad Sci U S A.

2011;108(29):12113-8. PMCID: 3141932.

176. Kim S, Burette A, Chung HS, Kwon SK, Woo J, Lee HW, et al. NGL family

PSD-95-interacting adhesion molecules regulate excitatory synapse formation. Nat

Neurosci. 2006;9(10):1294-301.

177. Woo J, Kwon SK, Choi S, Kim S, Lee JR, Dunah AW, et al. Trans-synaptic

adhesion between NGL-3 and LAR regulates the formation of excitatory synapses. Nat

Neurosci. 2009;12(4):428-37.

178. Song YS, Lee HJ, Prosselkov P, Itohara S, Kim E. Trans-induced cis interaction

in the tripartite NGL-1, netrin-G1 and LAR adhesion complex promotes development of

excitatory synapses. J Cell Sci. 2013;126(Pt 21):4926-38.

156

179. Hata Y, Butz S, Sudhof TC. CASK: a novel dlg/PSD95 homolog with an N-

terminal calmodulin-dependent protein kinase domain identified by interaction with

neurexins. J Neurosci. 1996;16(8):2488-94.

180. Cohen AR, Woods DF, Marfatia SM, Walther Z, Chishti AH, Anderson JM.

Human CASK/LIN-2 binds syndecan-2 and protein 4.1 and localizes to the basolateral

membrane of epithelial cells. J Cell Biol. 1998;142(1):129-38. PMCID: 2133028.

181. Hsueh YP, Yang FC, Kharazia V, Naisbitt S, Cohen AR, Weinberg RJ, et al.

Direct interaction of CASK/LIN-2 and syndecan heparan sulfate proteoglycan and their

overlapping distribution in neuronal synapses. J Cell Biol. 1998;142(1):139-51. PMCID:

2133027.

182. Biederer T, Sudhof TC. CASK and protein 4.1 support F-actin nucleation on

neurexins. J Biol Chem. 2001;276(51):47869-76.

183. Mukherjee K, Sharma M, Urlaub H, Bourenkov GP, Jahn R, Sudhof TC, et al.

CASK Functions as a Mg2+-independent neurexin kinase. Cell. 2008;133(2):328-39.

PMCID: 3640377.

184. Butz S, Okamoto M, Sudhof TC. A tripartite protein complex with the potential to

couple synaptic vesicle exocytosis to cell adhesion in brain. Cell. 1998;94(6):773-82.

185. Biederer T, Sudhof TC. Mints as adaptors. Direct binding to neurexins and

recruitment of munc18. J Biol Chem. 2000;275(51):39803-6.

186. Samuels BA, Hsueh YP, Shu T, Liang H, Tseng HC, Hong CJ, et al. Cdk5

promotes synaptogenesis by regulating the subcellular distribution of the MAGUK family

member CASK. Neuron. 2007;56(5):823-37. PMCID: 2151975.

157

187. Grootjans JJ, Zimmermann P, Reekmans G, Smets A, Degeest G, Durr J, et al.

Syntenin, a PDZ protein that binds syndecan cytoplasmic domains. Proc Natl Acad Sci U

S A. 1997;94(25):13683-8. PMCID: 28366.

188. Grootjans JJ, Reekmans G, Ceulemans H, David G. Syntenin-syndecan binding

requires syndecan-synteny and the co-operation of both PDZ domains of syntenin. J Biol

Chem. 2000;275(26):19933-41.

189. Ko J, Yoon C, Piccoli G, Chung HS, Kim K, Lee JR, et al. Organization of the

presynaptic active zone by ERC2/CAST1-dependent clustering of the tandem PDZ

protein syntenin-1. J Neurosci. 2006;26(3):963-70.

190. Takao-Rikitsu E, Mochida S, Inoue E, Deguchi-Tawarada M, Inoue M, Ohtsuka

T, et al. Physical and functional interaction of the active zone proteins, CAST, RIM1, and

Bassoon, in neurotransmitter release. J Cell Biol. 2004;164(2):301-11. PMCID: 2172332.

191. Ko J, Na M, Kim S, Lee JR, Kim E. Interaction of the ERC family of RIM-

binding proteins with the liprin-alpha family of multidomain proteins. J Biol Chem.

2003;278(43):42377-85.

192. Ohtsuka T, Takao-Rikitsu E, Inoue E, Inoue M, Takeuchi M, Matsubara K, et al.

Cast: a novel protein of the cytomatrix at the active zone of synapses that forms a ternary

complex with RIM1 and munc13-1. J Cell Biol. 2002;158(3):577-90. PMCID: 2173811.

193. Waites CL, Leal-Ortiz SA, Andlauer TF, Sigrist SJ, Garner CC. Piccolo regulates

the dynamic assembly of presynaptic F-actin. J Neurosci. 2011;31(40):14250-63.

PMCID: 3210199.

158

194. Olsen O, Moore KA, Fukata M, Kazuta T, Trinidad JC, Kauer FW, et al.

Neurotransmitter release regulated by a MALS-liprin-alpha presynaptic complex. J Cell

Biol. 2005;170(7):1127-34. PMCID: 2171538.

195. Zhen M, Jin Y. The liprin protein SYD-2 regulates the differentiation of

presynaptic termini in C. elegans. Nature. 1999;401(6751):371-5.

196. Dai Y, Taru H, Deken SL, Grill B, Ackley B, Nonet ML, et al. SYD-2 Liprin-

alpha organizes presynaptic active zone formation through ELKS. Nat Neurosci.

2006;9(12):1479-87.

197. Patel MR, Lehrman EK, Poon VY, Crump JG, Zhen M, Bargmann CI, et al.

Hierarchical assembly of presynaptic components in defined C. elegans synapses. Nat

Neurosci. 2006;9(12):1488-98. PMCID: 3917495.

198. Kaufmann N, DeProto J, Ranjan R, Wan H, Van Vactor D. Drosophila liprin-

alpha and the receptor phosphatase Dlar control synapse morphogenesis. Neuron.

2002;34(1):27-38.

199. Kittelmann M, Hegermann J, Goncharov A, Taru H, Ellisman MH, Richmond JE,

et al. Liprin-alpha/SYD-2 determines the size of dense projections in presynaptic active

zones in C. elegans. J Cell Biol. 2013;203(5):849-63. PMCID: 3857474.

200. Spangler SA, Schmitz SK, Kevenaar JT, de Graaff E, de Wit H, Demmers J, et al.

Liprin-alpha2 promotes the presynaptic recruitment and turnover of RIM1/CASK to

facilitate synaptic transmission. J Cell Biol. 2013;201(6):915-28. PMCID: 3678157.

201. Stigloher C, Zhan H, Zhen M, Richmond J, Bessereau JL. The presynaptic dense

projection of the Caenorhabditis elegans cholinergic neuromuscular junction localizes

159

synaptic vesicles at the active zone through SYD-2/liprin and UNC-10/RIM-dependent

interactions. J Neurosci. 2011;31(12):4388-96. PMCID: 3077722.

202. Wei Z, Zheng S, Spangler SA, Yu C, Hoogenraad CC, Zhang M. Liprin-mediated

large signaling complex organization revealed by the liprin-alpha/CASK and liprin-

alpha/liprin-beta complex structures. Mol Cell. 2011;43(4):586-98.

203. Holbrook S, Finley JK, Lyons EL, Herman TG. Loss of syd-1 from R7 neurons

disrupts two distinct phases of presynaptic development. J Neurosci. 2012;32(50):18101-

11. PMCID: 3663586.

204. Shipman SL, Nicoll RA. Dimerization of postsynaptic neuroligin drives synaptic

assembly via transsynaptic clustering of neurexin. Proc Natl Acad Sci U S A.

2012;109(47):19432-7. PMCID: 3511127.

205. Tanaka H, Miyazaki N, Matoba K, Nogi T, Iwasaki K, Takagi J. Higher-order

architecture of cell adhesion mediated by polymorphic synaptic adhesion molecules

neurexin and neuroligin. Cell Rep. 2012;2(1):101-10.

206. Wentzel C, Sommer JE, Nair R, Stiefvater A, Sibarita JB, Scheiffele P. mSYD1A,

a mammalian synapse-defective-1 protein, regulates synaptogenic signaling and vesicle

docking. Neuron. 2013;78(6):1012-23. PMCID: 3719036.

207. Washbourne P, Bennett JE, McAllister AK. Rapid recruitment of NMDA receptor

transport packets to nascent synapses. Nat Neurosci. 2002;5(8):751-9.

208. Niell CM, Meyer MP, Smith SJ. In vivo imaging of synapse formation on a

growing dendritic arbor. Nat Neurosci. 2004;7(3):254-60.

209. Ruthazer ES, Li J, Cline HT. Stabilization of axon branch dynamics by synaptic

maturation. J Neurosci. 2006;26(13):3594-603.

160

210. Kaether C, Skehel P, Dotti CG. Axonal membrane proteins are transported in

distinct carriers: a two-color video microscopy study in cultured hippocampal neurons.

Mol Biol Cell. 2000;11(4):1213-24.

211. Dresbach T, Hempelmann A, Spilker C, tom Dieck S, Altrock WD, Zuschratter

W, et al. Functional regions of the presynaptic cytomatrix protein bassoon: significance

for synaptic targeting and cytomatrix anchoring. Mol Cell Neurosci. 2003;23(2):279-91.

212. Hua Y, Sinha R, Martineau M, Kahms M, Klingauf J. A common origin of

synaptic vesicles undergoing evoked and spontaneous fusion. Nat Neurosci.13(12):1451-

3.

213. de Wit J, Toonen RF, Verhage M. Matrix-dependent local retention of secretory

vesicle cargo in cortical neurons. J Neurosci. 2009;29(1):23-37.

214. Linhoff MW, Lauren J, Cassidy RM, Dobie FA, Takahashi H, Nygaard HB, et al.

An unbiased expression screen for synaptogenic proteins identifies the LRRTM protein

family as synaptic organizers. Neuron. 2009;61(5):734-49. PMCID: 2746109.

215. Ripley B, Otto S, Tiglio K, Williams ME, Ghosh A. Regulation of synaptic

stability by AMPA receptor reverse signaling. Proc Natl Acad Sci U S A.

2011;108(1):367-72. PMCID: 3017209.

216. Stan A, Pielarski KN, Brigadski T, Wittenmayer N, Fedorchenko O, Gohla A, et

al. Essential cooperation of N-cadherin and neuroligin-1 in the transsynaptic control of

vesicle accumulation. Proc Natl Acad Sci U S A. 2010;107(24):11116-21. PMCID:

2890764.

217. Gokce O, Sudhof TC. Membrane-tethered monomeric neurexin LNS-domain

triggers synapse formation. J Neurosci. 2013;33(36):14617-28. PMCID: 3761060.

161

218. Tsuriel S, Geva R, Zamorano P, Dresbach T, Boeckers T, Gundelfinger ED, et al.

Local sharing as a predominant determinant of synaptic matrix molecular dynamics.

PLoS Biol. 2006;4(9):e271. PMCID: 1540708.

219. Kalla S, Stern M, Basu J, Varoqueaux F, Reim K, Rosenmund C, et al. Molecular

dynamics of a presynaptic active zone protein studied in Munc13-1-enhanced yellow

fluorescent protein knock-in mutant mice. J Neurosci. 2006;26(50):13054-66.

220. Tsuriel S, Fisher A, Wittenmayer N, Dresbach T, Garner CC, Ziv NE. Exchange

and redistribution dynamics of the cytoskeleton of the active zone molecule bassoon. J

Neurosci. 2009;29(2):351-8. PMCID: 2768361.

221. Fisher-Lavie A, Zeidan A, Stern M, Garner CC, Ziv NE. Use dependence of

presynaptic tenacity. J Neurosci. 2011;31(46):16770-80.

222. Staras K, Branco T. Sharing vesicles between central presynaptic terminals:

implications for synaptic function. Front Synaptic Neurosci. 2010;2:20. PMCID:

3059672.

223. Darcy KJ, Staras K, Collinson LM, Goda Y. Constitutive sharing of recycling

synaptic vesicles between presynaptic boutons. Nat Neurosci. 2006;9(3):315-21.

224. Zeidan A, Ziv NE. Neuroligin-1 loss is associated with reduced tenacity of

excitatory synapses. PLoS One. 2012;7(7):e42314. PMCID: 3409177.

225. Herzog E, Nadrigny F, Silm K, Biesemann C, Helling I, Bersot T, et al. In vivo

imaging of intersynaptic vesicle exchange using VGLUT1 Venus knock-in mice. J

Neurosci. 2011;31(43):15544-59.

162

226. LaMonte BH, Wallace KE, Holloway BA, Shelly SS, Ascano J, Tokito M, et al.

Disruption of dynein/dynactin inhibits axonal transport in motor neurons causing late-

onset progressive degeneration. Neuron. 2002;34(5):715-27.

227. Scott DA, Das U, Tang Y, Roy S. Mechanistic logic underlying the axonal

transport of cytosolic proteins. Neuron. 2011;70(3):441-54. PMCID: 3096075.

228. Krueger DD, Tuffy LP, Papadopoulos T, Brose N. The role of neurexins and

neuroligins in the formation, maturation, and function of vertebrate synapses. Curr Opin

Neurobiol. 2012;22(3):412-22.

229. Peixoto RT, Kunz PA, Kwon H, Mabb AM, Sabatini BL, Philpot BD, et al.

Transsynaptic signaling by activity-dependent cleavage of neuroligin-1.

Neuron.76(2):396-409. PMCID: PMC3783515.

230. Graf ER, Zhang X, Jin SX, Linhoff MW, Craig AM. Neurexins induce

differentiation of GABA and glutamate postsynaptic specializations via neuroligins. Cell.

2004;119(7):1013-26.

231. Biermann B, Sokoll S, Klueva J, Missler M, Wiegert JS, Sibarita JB, et al.

Imaging of molecular surface dynamics in brain slices using single-particle tracking. Nat

Commun.5:3024. PMCID: PMC3905702.

232. Fairless R, Masius H, Rohlmann A, Heupel K, Ahmad M, Reissner C, et al.

Polarized targeting of neurexins to synapses is regulated by their C-terminal sequences. J

Neurosci. 2008;28(48):12969-81.

233. Aiga M, Levinson JN, Bamji SX. N-cadherin and neuroligins cooperate to

regulate synapse formation in hippocampal cultures. J Biol Chem. 2011;286(1):851-8.

PMCID: 3013044.

163

234. Zhang W, Benson DL. Stages of synapse development defined by dependence on

F-actin. J Neurosci. 2001;21(14):5169-81.

235. Sun Y, Bamji SX. beta-Pix modulates actin-mediated recruitment of synaptic

vesicles to synapses. J Neurosci. 2011;31(47):17123-33.

236. Chia PH, Chen B, Li P, Rosen MK, Shen K. Local F-actin network links synapse

formation and axon branching. Cell. 2014;156(1-2):208-20. PMCID: 3954643.

237. Riedl J, Crevenna AH, Kessenbrock K, Yu JH, Neukirchen D, Bista M, et al.

Lifeact: a versatile marker to visualize F-actin. Nat Methods. 2008;5(7):605-7. PMCID:

2814344.

238. Smyth JW, Vogan JM, Buch PJ, Zhang SS, Fong TS, Hong TT, et al. Actin

cytoskeleton rest stops regulate anterograde traffic of connexin 43 vesicles to the plasma

membrane. Circ Res. 2012;110(7):978-89. PMCID: 3621031.

239. Flannery RJ, Bruses JL. N-cadherin induces partial differentiation of cholinergic

presynaptic terminals in heterologous cultures of brainstem neurons and CHO cells. Front

Synaptic Neurosci. 2012;4:6. PMCID: 3514636.

240. Shcherbakova DM, Verkhusha VV. Near-infrared fluorescent proteins for

multicolor in vivo imaging. Nat Methods. 2013;10(8):751-4. PMCID: 3737237.

241. Cantallops I, Cline HT. Synapse formation: if it looks like a duck and quacks like

a duck. Curr Biol. 2000;10(17):R620-3.

242. Craig AM, Kang Y. Neurexin-neuroligin signaling in synapse development. Curr

Opin Neurobiol. 2007;17(1):43-52. PMCID: PMC2820508.

164

243. Wienisch M, Klingauf J. Vesicular proteins exocytosed and subsequently

retrieved by compensatory endocytosis are nonidentical. Nat Neurosci. 2006;9(8):1019-

27.

244. Hoy JL, Constable JR, Vicini S, Fu Z, Washbourne P. SynCAM1 recruits NMDA

receptors via protein 4.1B. Mol Cell Neurosci. 2009;42(4):466-83. PMCID: 2784006.

245. Hirao K, Hata Y, Ide N, Takeuchi M, Irie M, Yao I, et al. A novel multiple PDZ

domain-containing molecule interacting with N-methyl-D-aspartate receptors and

neuronal cell adhesion proteins. J Biol Chem. 1998;273(33):21105-10.

246. Fame RM, MacDonald JL, Macklis JD. Development, specification, and diversity

of callosal projection neurons. Trends Neurosci. 2011;34(1):41-50. PMCID: 3053014.

247. Donahoo AL, Richards LJ. Understanding the mechanisms of callosal

development through the use of transgenic mouse models. Semin Pediatr Neurol.

2009;16(3):127-42.

248. Lohmann C, Kessels HW. The developmental stages of synaptic plasticity. J

Physiol. 2014;592(Pt 1):13-31. PMCID: 3903349.

249. Clement JP, Aceti M, Creson TK, Ozkan ED, Shi Y, Reish NJ, et al. Pathogenic

SYNGAP1 mutations impair cognitive development by disrupting maturation of dendritic

spine synapses. Cell. 2012;151(4):709-23. PMCID: 3500766.

250. Sytnyk V, Leshchyns'ka I, Delling M, Dityateva G, Dityatev A, Schachner M.

Neural cell adhesion molecule promotes accumulation of TGN organelles at sites of

neuron-to-neuron contacts. J Cell Biol. 2002;159(4):649-61.

251. Lee KD, Hollenbeck PJ. Phosphorylation of kinesin in vivo correlates with

organelle association and neurite outgrowth. J Biol Chem. 1995;270(10):5600-5.

165

252. Hollenbeck PJ. Phosphorylation of neuronal kinesin heavy and light chains in

vivo. J Neurochem. 1993;60(6):2265-75.

253. Hamasaki T, Barkalow K, Richmond J, Satir P. cAMP-stimulated

phosphorylation of an axonemal polypeptide that copurifies with the 22S dynein arm

regulates microtubule translocation velocity and swimming speed in Paramecium. Proc

Natl Acad Sci U S A. 1991;88(18):7918-22. PMCID: 52416.

254. Dillman JF, 3rd, Pfister KK. Differential phosphorylation in vivo of cytoplasmic

dynein associated with anterogradely moving organelles. J Cell Biol. 1994;127(6 Pt

1):1671-81. PMCID: 2120311.

255. Itzen A, Goody RS. GTPases involved in vesicular trafficking: structures and

mechanisms. Semin Cell Dev Biol. 2011;22(1):48-56.

256. Segev N. GTPases in intracellular trafficking: an overview. Semin Cell Dev Biol.

2011;22(1):1-2.

257. Chavrier P, Goud B. The role of ARF and Rab GTPases in membrane transport.

Curr Opin Cell Biol. 1999;11(4):466-75.

258. Bamji SX, Rico B, Kimes N, Reichardt LF. BDNF mobilizes synaptic vesicles

and enhances synapse formation by disrupting cadherin-beta-catenin interactions. J Cell

Biol. 2006;174(2):289-99. PMCID: 2064188.

259. Goldstein AY, Wang X, Schwarz TL. Axonal transport and the delivery of pre-

synaptic components. Curr Opin Neurobiol. 2008;18(5):495-503. PMCID: 2653082.

260. Hirokawa N, Noda Y, Tanaka Y, Niwa S. Kinesin superfamily motor proteins and

intracellular transport. Nat Rev Mol Cell Biol. 2009;10(10):682-96.

166

261. Hamdan FF, Gauthier J, Araki Y, Lin DT, Yoshizawa Y, Higashi K, et al. Excess

of de novo deleterious mutations in genes associated with glutamatergic systems in

nonsyndromic intellectual disability. Am J Hum Genet. 2011;88(3):306-16. PMCID:

3059427.

262. Hopf FW, Waters J, Mehta S, Smith SJ. Stability and plasticity of developing

synapses in hippocampal neuronal cultures. J Neurosci. 2002;22(3):775-81.

263. Murthy VN, Schikorski T, Stevens CF, Zhu Y. Inactivity produces increases in

neurotransmitter release and synapse size. Neuron. 2001;32(4):673-82.

264. Matz J, Gilyan A, Kolar A, McCarvill T, Krueger SR. Rapid structural alterations

of the active zone lead to sustained changes in neurotransmitter release. Proc Natl Acad

Sci U S A. 2010;107(19):8836-41. PMCID: 2889309.

265. Holderith N, Lorincz A, Katona G, Rozsa B, Kulik A, Watanabe M, et al. Release

probability of hippocampal glutamatergic terminals scales with the size of the active

zone. Nat Neurosci. 2012;15(7):988-97. PMCID: 3386897.

266. Minerbi A, Kahana R, Goldfeld L, Kaufman M, Marom S, Ziv NE. Long-term

relationships between synaptic tenacity, synaptic remodeling, and network activity. PLoS

Biol. 2009;7(6):e1000136. PMCID: 2693930.

267. Staras K, Branco T, Burden JJ, Pozo K, Darcy K, Marra V, et al. A vesicle

superpool spans multiple presynaptic terminals in hippocampal neurons. Neuron.

2010;66(1):37-44. PMCID: 2908741.

268. Fremeau RT, Jr., Kam K, Qureshi T, Johnson J, Copenhagen DR, Storm-Mathisen

J, et al. Vesicular glutamate transporters 1 and 2 target to functionally distinct synaptic

release sites. Science. 2004;304(5678):1815-9.

167

269. Berry CT, Sceniak MP, Zhou L, Sabo SL. Developmental up-regulation of

vesicular glutamate transporter-1 promotes neocortical presynaptic terminal development.

PLoS One. 2012;7(11):e50911. PMCID: 3511412.

270. Sceniak MP, Berry CT, Sabo SL. Facilitation of neocortical presynaptic terminal

development by NMDA receptor activation. Neural Dev. 2012;7:8. PMCID: 3296626.

271. Kwon HB, Sabatini BL. Glutamate induces de novo growth of functional spines

in developing cortex. Nature.474(7349):100-4. PMCID: PMC3107907.

272. Mozhayeva MG, Sara Y, Liu X, Kavalali ET. Development of vesicle pools

during maturation of hippocampal synapses. J Neurosci. 2002;22(3):654-65.

273. Owald D, Fouquet W, Schmidt M, Wichmann C, Mertel S, Depner H, et al. A

Syd-1 homologue regulates pre- and postsynaptic maturation in Drosophila. J Cell Biol.

2010;188(4):565-79. PMCID: PMC2828917.

274. Spangler SA, Hoogenraad CC. Liprin-alpha proteins: scaffold molecules for

synapse maturation. Biochem Soc Trans. 2007;35(Pt 5):1278-82.

275. Chia PH, Patel MR, Shen K. NAB-1 instructs synapse assembly by linking

adhesion molecules and F-actin to active zone proteins. Nat Neurosci. 2012;15(2):234-

42. PMCID: 3848868.

276. Martinez A, Alcantara S, Borrell V, Del Rio JA, Blasi J, Otal R, et al. TrkB and

TrkC signaling are required for maturation and synaptogenesis of hippocampal

connections. J Neurosci. 1998;18(18):7336-50.

277. Collin C, Vicario-Abejon C, Rubio ME, Wenthold RJ, McKay RD, Segal M.

Neurotrophins act at presynaptic terminals to activate synapses among cultured

hippocampal neurons. Eur J Neurosci. 2001;13(7):1273-82.

168

278. Otal R, Martinez A, Soriano E. Lack of TrkB and TrkC signaling alters the

synaptogenesis and maturation of mossy fiber terminals in the hippocampus. Cell Tissue

Res. 2005;319(3):349-58.

279. Melo CV, Mele M, Curcio M, Comprido D, Silva CG, Duarte CB. BDNF

regulates the expression and distribution of vesicular glutamate transporters in cultured

hippocampal neurons. PLoS One. 2013;8(1):e53793. PMCID: 3543267.

280. Alsina B, Vu T, Cohen-Cory S. Visualizing synapse formation in arborizing optic

axons in vivo: dynamics and modulation by BDNF. Nat Neurosci. 2001;4(11):1093-101.

281. Andlauer TF, Sigrist SJ. In vivo imaging of Drosophila larval neuromuscular

junctions to study synapse assembly. Cold Spring Harb Protoc. 2012;2012(4):407-13.

282. Hu B, Nikolakopoulou AM, Cohen-Cory S. BDNF stabilizes synapses and

maintains the structural complexity of optic axons in vivo. Development.

2005;132(19):4285-98.

283. Jontes JD, Emond MR, Smith SJ. In vivo trafficking and targeting of N-cadherin

to nascent presynaptic terminals. J Neurosci. 2004;24(41):9027-34.

284. Maeder CI, San-Miguel A, Wu EY, Lu H, Shen K. In vivo neuron-wide analysis

of synaptic vesicle precursor trafficking. Traffic. 2014;15(3):273-91.

285. Meyer MP, Smith SJ. Evidence from in vivo imaging that synaptogenesis guides

the growth and branching of axonal arbors by two distinct mechanisms. J Neurosci.

2006;26(13):3604-14.

286. Panzer JA, Song Y, Balice-Gordon RJ. In vivo imaging of preferential motor

axon outgrowth to and synaptogenesis at prepatterned acetylcholine receptor clusters in

embryonic zebrafish skeletal muscle. J Neurosci. 2006;26(3):934-47.

169

287. Turney SG, Walsh MK, Lichtman JW. In vivo imaging of the developing

neuromuscular junction in neonatal mice. Cold Spring Harb Protoc. 2012;2012(11):1166-

76.

288. Zhou HM, Brust-Mascher I, Scholey JM. Direct visualization of the movement of

the monomeric axonal transport motor UNC-104 along neuronal processes in living

Caenorhabditis elegans. J Neurosci. 2001;21(11):3749-55.

289. Levitan ES, Lanni F, Shakiryanova D. In vivo imaging of vesicle motion and

release at the Drosophila neuromuscular junction. Nat Protoc. 2007;2(5):1117-25.

290. Lendvai B, Stern EA, Chen B, Svoboda K. Experience-dependent plasticity of

dendritic spines in the developing rat barrel cortex in vivo. Nature. 2000;404(6780):876-

81.

291. Trachtenberg JT, Chen BE, Knott GW, Feng G, Sanes JR, Welker E, et al. Long-

term in vivo imaging of experience-dependent synaptic plasticity in adult cortex. Nature.

2002;420(6917):788-94.

292. Holtmaat AJ, Trachtenberg JT, Wilbrecht L, Shepherd GM, Zhang X, Knott GW,

et al. Transient and persistent dendritic spines in the neocortex in vivo. Neuron.

2005;45(2):279-91.

293. De Paola V, Holtmaat A, Knott G, Song S, Wilbrecht L, Caroni P, et al. Cell type-

specific structural plasticity of axonal branches and boutons in the adult neocortex.

Neuron. 2006;49(6):861-75.

294. Stettler DD, Yamahachi H, Li W, Denk W, Gilbert CD. Axons and synaptic

boutons are highly dynamic in adult visual cortex. Neuron. 2006;49(6):877-87.

170

295. Majewska AK, Newton JR, Sur M. Remodeling of synaptic structure in sensory

cortical areas in vivo. J Neurosci. 2006;26(11):3021-9.

296. Yamahachi H, Marik SA, McManus JN, Denk W, Gilbert CD. Rapid axonal

sprouting and pruning accompany functional reorganization in primary visual cortex.

Neuron. 2009;64(5):719-29. PMCID: 2818836.

297. Holtmaat A, Bonhoeffer T, Chow DK, Chuckowree J, De Paola V, Hofer SB, et

al. Long-term, high-resolution imaging in the mouse neocortex through a chronic cranial

window. Nat Protoc. 2009;4(8):1128-44. PMCID: 3072839.

298. Goldey GJ, Roumis DK, Glickfeld LL, Kerlin AM, Reid RC, Bonin V, et al.

Removable cranial windows for long-term imaging in awake mice. Nat Protoc.

2014;9(11):2515-38. PMCID: 4442707.

299. Iida J, Hirabayashi S, Sato Y, Hata Y. Synaptic scaffolding molecule is involved

in the synaptic clustering of neuroligin. Mol Cell Neurosci. 2004;27(4):497-508.

300. Um JW, Ko J. LAR-RPTPs: synaptic adhesion molecules that shape synapse

development. Trends Cell Biol. 2013;23(10):465-75.

301. Dunah AW, Hueske E, Wyszynski M, Hoogenraad CC, Jaworski J, Pak DT, et al.

LAR receptor protein tyrosine phosphatases in the development and maintenance of

excitatory synapses. Nat Neurosci. 2005;8(4):458-67.

302. Siu R, Fladd C, Rotin D. N-cadherin is an in vivo substrate for protein tyrosine

phosphatase sigma (PTPsigma) and participates in PTPsigma-mediated inhibition of axon

growth. Mol Cell Biol. 2007;27(1):208-19. PMCID: 1800655.

171

303. Besse F, Mertel S, Kittel RJ, Wichmann C, Rasse TM, Sigrist SJ, et al. The Ig cell

adhesion molecule Basigin controls compartmentalization and vesicle release at

Drosophila melanogaster synapses. J Cell Biol. 2007;177(5):843-55. PMCID: 2064284.

304. Stavoe AK, Colon-Ramos DA. Netrin instructs synaptic vesicle clustering

through Rac GTPase, MIG-10, and the actin cytoskeleton. J Cell Biol. 2012;197(1):75-

88. PMCID: 3317799.

305. Goldman JS, Ashour MA, Magdesian MH, Tritsch NX, Harris SN, Christofi N, et

al. Netrin-1 promotes excitatory synaptogenesis between cortical neurons by initiating

synapse assembly. J Neurosci. 2013;33(44):17278-89.

306. Pawson C, Eaton BA, Davis GW. Formin-dependent synaptic growth: evidence

that Dlar signals via Diaphanous to modulate synaptic actin and dynamic pioneer

microtubules. J Neurosci. 2008;28(44):11111-23. PMCID: 2654187.

307. Pielage J, Bulat V, Zuchero JB, Fetter RD, Davis GW. Hts/Adducin controls

synaptic elaboration and elimination. Neuron. 2011;69(6):1114-31. PMCID: 3073818.

308. Coyle IP, Koh YH, Lee WC, Slind J, Fergestad T, Littleton JT, et al. Nervous

wreck, an SH3 adaptor protein that interacts with Wsp, regulates synaptic growth in

Drosophila. Neuron. 2004;41(4):521-34.

309. Rodal AA, Motola-Barnes RN, Littleton JT. Nervous wreck and Cdc42 cooperate

to regulate endocytic actin assembly during synaptic growth. J Neurosci.

2008;28(33):8316-25. PMCID: 2546611.

310. Gross GG, Junge JA, Mora RJ, Kwon HB, Olson CA, Takahashi TT, et al.

Recombinant probes for visualizing endogenous synaptic proteins in living neurons.

Neuron. 2013;78(6):971-85. PMCID: 3779638.

172

311. Nieland TJ, Logan DJ, Saulnier J, Lam D, Johnson C, Root DE, et al. High

content image analysis identifies novel regulators of synaptogenesis in a high-throughput

RNAi screen of primary neurons. PLoS One. 2014;9(3):e91744. PMCID: 3954765.

312. Sharma K, Choi SY, Zhang Y, Nieland TJ, Long S, Li M, et al. High-throughput

genetic screen for synaptogenic factors: identification of LRP6 as critical for excitatory

synapse development. Cell Rep. 2013;5(5):1330-41. PMCID: 3924421.

313. Niedringhaus M, Dumitru R, Mabb AM, Wang Y, Philpot BD, Allbritton NL, et

al. Transferable neuronal mini-cultures to accelerate screening in primary and induced

pluripotent stem cell-derived neurons. Sci Rep. 2015;5:8353. PMCID: 4322355.

314. Kim JE, O'Sullivan ML, Sanchez CA, Hwang M, Israel MA, Brennand K, et al.

Investigating synapse formation and function using human pluripotent stem cell-derived

neurons. Proc Natl Acad Sci U S A. 2011;108(7):3005-10. PMCID: 3041068.

315. Lancaster MA, Renner M, Martin CA, Wenzel D, Bicknell LS, Hurles ME, et al.

Cerebral organoids model human brain development and microcephaly. Nature.

2013;501(7467):373-9. PMCID: 3817409.

316. Kadoshima T, Sakaguchi H, Nakano T, Soen M, Ando S, Eiraku M, et al. Self-

organization of axial polarity, inside-out layer pattern, and species-specific progenitor

dynamics in human ES cell-derived neocortex. Proc Natl Acad Sci U S A.

2013;110(50):20284-9. PMCID: 3864329.