POSTSYNAPTIC EFFECTORS OF NEURON MORPHOLOGY AND FUNCTION: PART I. CHARACTERIZATION OF POSTSYNAPTIC DROSOPHILA SYNDAPIN PART II. CHIMERIC LIGHT-ACTIVATED RECEPTORS
FOR THE CONTROL OF 5-HT1A SIGNALING
by
EUGENE OH
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy
Dissertation advisor: Dr. Stefan Herlitze
Department of Neurosciences CASE WESTERN RESERVE UNIVERSITY January 2011
CASE WESTERN RESERVE UNIVERSITY SCHOOL OF GRADUATE STUDIES
We hereby approve the thesis/dissertation of
Eugene Oh
candidate for the Doctor of Philosophy degree *.
(signed) Dr. Gary Landreth (chair of the committee)
Dr. Stefan Herlitze
Dr. Jerry Silver
Dr. Stephen Maricich
(date) November 19th, 2010
*We also certify that written approval has been obtained for any proprietary material contained therein.
Copyright © 2011 by Eugene Oh All rights reserved
DEDICATION
This work is dedicated to my friends and
family for their enduring support and
encouragement; my children Lucia, Paloma
and Elliot who inspire me to be greater
person; and most importantly, to my wife
Sarah. Nothing would have been possible
without her help and unconditional love.
iv TABLE OF CONTENTS
Title Page ...... i
Signature page ...... ii
Dedication ...... iv
Table of Contents ...... v
List of Figures ...... ix
Acknowledgements ...... xi
Abstract ...... xii
CHAPTER 1: INTRODUCTION TO PART I ...... 1
The genetic tractability of Drosophila melanogaster ...... 2
The Drosophila neuromuscular junction (NMJ) as a model for synaptic plasticity ...... 6
Regulation of Drosophila NMJ morphology ...... 9
Transsynaptic signaling regulates synaptic growth: ...... 9
Actin cytoskeleton dependent plasticity ...... 11
The function of Drosophila Wsp and linking actin to NMJ growth ...... 15
Endocytic proteins are negative regulators of NMJ growth ...... 17
Vertebrate Syndapin ...... 20
F-BAR Domain containing proteins: structural bridges between cytoskeleton and cell membranes ...... 22
Drosophila Syndapin ...... 24
Research aims ...... 26
v CHAPTER 2: DROSOPHILA SYNDAPIN INTERACTS WITH WISKOTT-ALRICH SYNDROME PROTEIN AND REGULATES NEUROMUSCULAR JUNCTION GROWTH POSTSYNAPTICALLY ...... 28
Introduction ...... 29
Materials and Methods ...... 32
Results ...... 36
Identification of a Drosophila Syndapin homologue...... 36
Identification P-element inserts near Drosophila Syndapin and characterization of Syndapin hypomorphic alleles ...... 37
Localization of Syndapin at Drosophila neuromuscular junctions...... 39
Loss of Syndapin results in synaptic overgrowth ...... 40
Postsynaptic expression of a Synd transgene rescues the overgrowth phenotype ...... 42
Drosophila Syndapin binds Wsp and the conserved proline487 residue of residue of Synd is required for this interaction ...... 43
Syndapin and Wsp interact in vivo ...... 44
Synd acts upstream of wsp ...... 45
Figures ...... 47
CHAPTER 3: DISCUSSION ...... 63
Research Conclusions ...... 64
Postsynaptic actin cytoskeleton and the control of synaptic morphology ...... 64
Syndapin is an adaptor protein that recruits Wsp to the postsynaptic membrane ...... 66
A putative role for Drosophila Syndapin in synaptic vesicle endocytosis ...... 67
Remaining questions and future directions ...... 68
vi CHAPTER 4: INTRODUCTION TO PART II ...... 71
The function and importance of the serotonergic system ...... 72
The 5-HT1A receptor and its role in disease pathogenesis and treatment ...... 75
Structural determinants of 5-HT receptor function and targeting ...... 79
G protein-coupled receptors of visual systems as "optogenetic" probes ...... 81
Invertebrate rhodopsin ...... 83
Vertebrate rhodopsins ...... 84
Melanopsin ...... 90
OptoXRs: chimeric GPCRs for the light-based control of specific GPCR intracellular signaling pathways and behavior ...... 91
Research goals ...... 94
CHAPTER 5: CONTROL OF 5-HT1A RECEPTOR SIGNALING BY LIGHT-ACTIVATED G PROTEIN-COUPLED RECEPTORS ...... 96
Introduction ...... 97
Materials and Methods ...... 100
Results ...... 108
Cloning of Rh-CT5-HT1A and optimization of the light activation paradigm...... 108
Expression pattern and function of Rh-CT5-HT1A resembles fluorescently tagged 5-HT1A in HEK cells...... 111
Subcellular targeting of Rh-CT5-HT1A resembles that of 5-HT1A...... 114
Rh-CT5-HT1A is functional in cultured hippocampal neurons and competitively inhibits endogenous 5-HT1A receptor...... 116
Rh-CT5-HT1A compensates for the loss of 5-HT1A signaling in cultured hippocampal neurons of 5-HT1A null mice...... 118
Rh-CT5-HT1A functionally substitutes for 5-HT1A signaling in dorsal
vii raphe nucleus neurons in brain slices from 5-HT1A null mice...... 119
Figures ...... 121
CHAPTER 6: DISCUSSION ...... 146
Research conclusions ...... 147
Critical trafficking domains as molecular tags to direct intracellular targeting ...... 148
Endogenous receptor replacement by exogenous receptor expression ...... 149
Remaining questions and future directions ...... 150
REFERENCES ...... 156
viii LIST OF FIGURES
FIGURES FOR PART I
Figure 1 - Amino acid sequence alignment of Drosophila Synd with other known Synd isoforms ...... 47
Figure 2 - Schematic representation of the gene structure of Drosophila Synd. .. 49
Figure 3 - In wild type embryos and larvae, Synd is distributed in actin-rich, highly curved structures...... 51
Figure 4 - Far western blots using GST-Synd-SH3 domains to bind recombinant Wsp...... 53
Figure 5 - Localization of Synd at the neuromuscular junction...... 55
Figure 6 - Loss of postsynaptic Synd leads to altered neuromuscular junction morphology ...... 57
Figure 7 - Gene dosage of Synd and Wsp affects hyperbranching phenotype .... 59
Figure 8 - Post-synaptic Wsp is decreased in a Synd mutant ...... 61
FIGURES FOR PART I
Figure 1 - Vertebrate rhodopsin does not activate GIRK channels in the absence of light stimulus...... 121
Figure 2 - Amino acid sequence alignments of mCherry tagged GPCRs Rh, 5-HT1A and Rh-CT5-HT1A ...... 123
Figure 3 - Protein sequence alignments of Rh, 5-HT1A-mCherry and OptoXRs ...... 125
Figure 4 - Opto-5-HT1A is functional in HEK293 cells, but exhibits atypical GPCR activation and fails to inactivate ...... 127
Figure 5 - Functional expression and characterization of Rh-CT5-HT1A in HEK293 cells...... 129
ix Figure 6 - Decremental decrease in GIRK channel activation by Rh in HEK cells with repeated light stimulation ...... 131
Figure 7 - The C-terminal domain of 5-HT1A is sufficient to induce targeting of vertebrate rhodopsin somatodendritically in neurons...... 133
Figure 8 - The C-terminal domain of 5-HT1A receptor promotes distal targeting within dendrites of hippocampal neurons...... 135
Figure 9 - Rh-CT5-HT1A induces membrane hyperpolarization in rat hippocampal neurons with light stimulus...... 137
Figure 10 - Rh-CT5-HT1A replaces endogenous 5-HT1A receptors in hippocampal neurons...... 139
Figure 11 - Light activation of Rh-CT5-HT1A functionally rescues 5-HT1A loss-of-function phenotype in cultured hippocampal neurons...... 141
Figure 12 - Rh-CT5-HT1A compensates for loss of 5-HT1A mediated signaling of neurons in the dorsal raphe nuclei of 5-HT1A KO mice...... 144
x ACKNOWLEDGEMENTS
I first have to thank my two thesis advisors for providing excellent training
opportunities and mentorship during my Ph.D. career. Dr. Iain Robinson mentored me
for Part I of this dissertation (Drosophila genetics and synaptic development) and Dr.
Stefan Herlitze for Part II (Light-activated chimeric receptor development). My fellow lab members have been instrumental for day to day advice and assistance with
experiments. I would also like to thank past and present members of my thesis
committee for helping to guide my work: Drs. Gary Landreth (Chair), Lynn Landmesser,
Jerry Silver, Stephan Maricich, Heather Broihier, Philip Morgan, and Bob Miller. I am
especially grateful to Dr. Gary Landreth and all the members of the Landreth lab for
twice providing me with a laboratory home. This work could not have been completed
without their help. I also want to thank Maryanne Pendergast for help with confocal
microscopy and image processing
I am grateful to several people who provided critical reagents: Dr. Eyal Schejter
for the wsp cDNA and Wsp antibody; Dr. Noreen Reist for the synaptotagmin antibody;
and Dr. Helen Salz for the anti-SNF antibody. I would like to thank Dr. Lawrence
Tecott for the 5-HT1A knock-out mice. Drs. Cahir O’Kane, Heather Broihier, Philip
Morgan, Gary Landreth, Bob Miller, and Stefan Herlitze provided very helpful discussion and critical reading of manuscripts.
My thesis work was supported by a NIH NRSA Predoctoral Fellowship (F30
MH084371), the MSTP Training Grant (T32 GM007250), and the Neurosciences
Training Grant (T32 AG000271). I was also supported by a NIH grant to Dr. Stefan
Herlitze (R01 MH081127) and a NIH SBIR to LucCell, Inc. (R43 MH083302).
xi
Postsynaptic Effectors of Neuron Morphology and Function: Part I. Characterization of Postsynaptic Drosophila Syndapin Part II. Chimeric Light-Activated Receptors for the Control of 5-HT1A Signaling
Abstract
by
EUGENE OH
PART I
Vertebrate syndapin (Synd) was identified as a dynamin interacting protein,
which binds Wiskott-Aldrich Syndrome protein (WASp). WASp is a known activator of
the Arp-2/3 complex (actin related protein) that promotes actin polymerization. In
Drosophila, Synd has a postsynaptic distribution at neuromuscular junctions (NMJ). To
determine the synaptic function of Synd, we have generated and characterized
Drosophila Synd loss-of-function mutants. These mutants display a NMJ overgrowth
phenotype manifested by increased synaptic span, bouton number, and extent of
branching. The synd mutants also contain hyperbranching boutons that give rise to
multiple "satellite" boutons. We can rescue this phenotype by expression of a synd
transgene postsynaptically, but not presynaptically. Loss of synd resembles the NMJ
morphological phenotype of wsp loss-of-function mutants, suggesting a common functional pathway. Furthermore, the NMJ overgrowth phenotype is most pronounced in synd and wsp double mutants and dependent on gene dosage--demonstrating that the two
xii genes interact. Finally, relative postsynaptic staining of Wsp is decreased in a synd mutant, but Synd remains intact in a wsp mutant. Taken together, this suggests that
Syndapin serves as an adaptor protein recruiting Wsp postsynaptically to the NMJ. Loss
of synd is likely to modulate synapse growth by disrupting Wsp localization and function.
PART II
The 5-HT1A receptor is a metabotropic G protein-coupled receptor (GPCR) linked
to the Gi/o signaling pathway and has been specifically implicated in the pathogenesis of depression and anxiety. To understand and precisely control 5-HT1A signaling, we
created a light-activated GPCR, which targets into 5-HT1A receptor domains and
substitutes for endogenous 5-HT1A receptors. To induce 5-HT1A-like targeting, vertebrate
rhodopsin (Rh) was tagged with the C-terminal domain (CT) of 5-HT1A (Rh-CT5-HT1A).
+ Rh-CT5-HT1A activates G protein-coupled inward rectifying K channels (GIRK) in
response to light and causes membrane hyperpolarization in hippocampal neurons,
similar to the agonist-induced responses of the 5-HT1A receptor. The intracellular
distribution of Rh-CT5-HT1A resembles that of wild type 5-HT1A receptor; Rh-CT5-HT1A
localizes somatodendritically and efficiently traffics to distal dendritic processes.
Additionally, neuronal expression of Rh-CT5-HT1A, but not Rh, inhibits responses to 5-
HT1A agonist, suggesting that Rh-CT5-HT1A and endogenous 5-HT1A receptors compete to
interact with the same trafficking machinery. Finally, Rh-CT5-HT1A is able to substitute
for absent 5-HT1A signaling in cultured neurons and dorsal raphe slices of KO mice,
showing that Rh-CT5-HT1A functionally compensates for native 5-HT1A. Thus, as an
optogenetic tool, Rh-CT5-HT1A has the potential to directly correlate in vivo 5-HT1A
xiii signaling with 5-HT neuron activity and behavior in both wild type animals and animal models of neuropsychiatric disease.
xiv
CHAPTER 1
INTRODUCTION TO PART I
1
The genetic tractability of Drosophila melanogaster
As a genetic model, Drosophila melanogaster is an attractive model system as it provides a relatively small genome that is easily manipulated and can be rendered invariant in stable lines. Once established, the maintenance of stable lines is minimal since genotyping is unnecessary and flies are allowed to reproduce freely in vials or bottles containing culture media. The Drosophila genome was fully sequence in 2000
and is still being refined (Adams et al., 2000; Lloyd et al., 2000). It consists of four
chromosomes (X/Y, 2, 3, 4) where chromosomes Y and four are short enough that
virtually no recombination occurs during meiosis (Greenspan, 2004). The use of
"balancers" eliminates viable homologous recombination events in the remaining
chromosomes. Balancer chromosomes contain multiple, nested inversions so that
crossing over between a balancer and its homologous chromosome during meiosis results
in a pair of defective chromatids (Greenspan, 2004). Many genes are lost whereas other
genes are duplicated for these recombined chromatids. Therefore, progeny carrying
chromosomes produced by recombination between normal chromosomes and balancers
are not viable. This reduces the complexity of inheritance in Drosophila strains to simple
Mendelian genetics.
Drosophila can be molecularly genotyped using PCR and sequencing techniques,
but generally, it is done phenotypically by the use of dominant, visible markers that are
easily distinguished such as changes in body pigmentation, eye color, bristle
number/shape, wing morphology, or transgenically expressed GFP. Good balancer
chromosomes carry easily distinguishable, highly penetrant markers that are the product
of mutations at the inversion breakpoints and/or carry transgenic modifications (i.e.
2
GFP). Furthermore, many of these marker mutations on balancers also cause recessive
lethality. The following would be an example of a cross using balancers, where the goal
is to generate a fly with the Ly allele (on 3rd chromosome) balanced by TM6B (3rd chromosome):
All surviving progeny of this cross would be Ly/TM6B, Ly/MKRS, or TM3/TM6B. The allelic combination of MKRS and TM3 would not be found since homozygous stubble
(Sb) mutation is homozygous lethal. Flies that lack anterior and posterior wing borders
(Lyra, Ly) and have extra humeral "shoulder" bristles (Hu), but have neither short, stubbly bristles (Sb) nor dark body pigmentation (ebony, e) would be selected. Ly/TM6B flies would not be ebony since they would be heterozygous for this recessive trait, denoted in lowercase. Since Ly and TM6B cause homozygous lethality, Ly/TM6B is a stable stock and all progeny of Ly/TM6B crossed to itself (F2) would be Ly/TM6B.
The construction and mapping of mutants and transgenic lines is rapid and straight-forward. Drosophila are well-suited for chemical/radiation mutagenesis since they can be grown large numbers and mutants can be rapidly identified, depending on the screening parameters. For the generation of cell-type specific transgenic lines, the most common approach is the Gal-4/UAS binary system (Brand and Perrimon, 1993). Plasmid
DNA containing upstream activating sequence (UAS) and gene of interest within a transposable element (such as P-element) is microinjected directly into fertilized embryos along with a helper plasmid encoding a transposase (Dahmann, 2008). The transposase induces stable integration of the transposon. Several distinct, balanced lines can be generated and mapped to chromosome within six weeks because of the short generation
3
time of Drosophila. These lines can then be crossed to transgenic flies expressing Gal-4 under the control of cell-type specific promoters. Transgenic lines are not only used to show misexpression/overexpression phenotypes, but are critical for the explicit demonstration of gene function by mutational analysis. Transgenic rescue of loss-of- function phenotypes remains the gold standard test of gene function validity.
Furthermore, the cell type(s) in which a particular gene functions can be determined by the use of cell-type specific Gal-4 lines. Indeed, similar strategies have been applied in vertebrates, but the ease of transgenesis combined with the several hundred characterized
Gal-4 drivers available for Drosophila greatly facilitates this approach (Duffy, 2002).
Genetic mapping is also relatively straightforward in Drosophila if the mutation/transgene also contains phenotypic markers. This is achieved by the use of balancer chromosomes to indentify the mutated chromosome. Then the use of marked chromosomes and sequencing are used to further refine genetic location. Unlike balancers, marked chromosomes do not contain inversions, but rather a series of phenotypic markers along the length of the entire chromosome. Crossing mutants to marked chromosome lines will result in varying recombinatorial events and progeny will retain different combinations of markers. The mutation can then be localized to a genomic region "between" two markers by tracking the mutant phenotype. Deficiency mapping is another alternative where mutants are subjected to complementation testing with a series of nested deficiency lines. The deficiency that uncovers the phenotype
(homozygous null) indicates map location. This type of mapping strategy of course depends on a easily identifiable, homozygous null phenotype (i.e. lethality). Subsequent sequencing will then allow precise mapping.
4
A major weakness of Drosophila genetics is the relative difficulty of generating
targeted mutations. Homologous recombination and other targeting strategies have been
developed, but because these approaches are relatively time consuming (several months -
a year) compared to other established mutation approaches, they are generally not
favored. Furthermore, just like in other species, targeted strategies can induce second site
mutations that can confound genetic analysis (Roy and Hart, 2010). Second site
mutational effects can be controlled for by the use of transheterozygotes (i.e. animal with
two independently derived mutant alleles of the same gene) instead of homozygous
mutants; still, this limits the utility of targeted mutations. Recently, there have been
technological advances incorporating the use of the attachment site for the phage ΦC31
integrase, attP, with homologous recombination (Gao et al., 2008). Subsequent to attP incorporation in the vicinity of the gene of interest, plasmids carrying attB attachment sites and the desired mutation(s) are incorporated by ΦC31 integrase. This type of strategy is also being applied to generated targeted transgenic lines (Bateman et al.,
2006). Despite the lack of convenient targeted mutational strategies until very recently, there are ever expanding libraries of mapped transposable element insertions (such as P- elements and PiggyBacs) that disrupt specific genes and deficiency lines (chromosomal deletions) such as the Berkeley Drosophila genome project (BDGP) and the Exelixis collection at Harvard Medical School (Bellen et al., 2004; Thibault et al., 2004). These transposon lines can be further manipulated to create additional mutants. Remobilization of the transposable elements can occasionally result in imprecise excisions leaving behind transposon DNA and/or creating genomic deletions (Hummel and Klambt, 2008).
5
Finally, Drosophila growth and development is temperature-dependent adding another degree of flexibility. Specifically, flies develop at half the rate at 18 °C as opposed to when they are cultured at 25 °C (Greenspan, 2004). Timed collections allows for the tight control of mating conditions by, for example, the unambiguous collection of virgin females so that subsequent crossing steps are genetically uncontaminated. The genetic tractability along with the short generation time (2 weeks per crossing step) makes Drosophila an ideal genetic model system for studying a wide range of biological processes.
The Drosophila neuromuscular junction (NMJ) as a model for synaptic plasticity
The Drosophila larval neuromuscular junction (NMJ) is a well established model
system for the study of synaptic development and function. It represents an intersection
between the powerful and elegant genetics of Drosophila, functional neuroscience and
embryonic neurodevelopment. NMJs are readily accessible synapses, hence, they are
amenable to a variety of techniques such as immunohistochemistry, electrophysiology,
Ca2+ imaging and electron microscopy. As a model organism, 50% of Drosophila genes
have Human homologs and approximately 3/4 of genes linked to disease have related
Drosophila homologs (Chintapalli et al., 2007; Reiter et al., 2001). Because Drosophila
NMJs are glutamatergic, they contain many of the same cellular and molecular components found in vertebrate excitatory synapses of the CNS (Collins and DiAntonio,
2007). Thus, the mechanisms that coordinate the development, maturation and maintenance of the Drosophila NMJ serve as promising candidates for regulating plasticity of glutamatergic synapses in general.
6
Wild type Drosophila larvae have a stereotyped body wall musculature pattern organized in repeating segments that are bilaterally symmetric. Each abdominal hemisegment has 30 individually identifiable muscles that are innervated by about 40 motor neurons (Gramates and Budnik, 1999). During late embryogenesis, NMJ synaptogenesis begins as motorneuron growth cones migrate out and extend filapodial processes in search for their target muscles (Gramates and Budnik, 1999). Once motorneurons make contact with their muscle targets, primitive, embryonic synapses are initially made and are continually modified during development. Coordinated maturation of both pre- (neuronal) and postsynaptic (muscle) aspects is needed to generate working synapses. Shortly after, synaptic varicosities called "boutons" begin to form along the length of the NMJ, each normally containing approximately 10 presynaptic active zones
(Collins and DiAntonio, 2007). The active zones represent neurotransmitter release sites where readily releasable clusters of synaptic vesicles; electron-dense structures called T- bars that most likely promote vesicular release; and Ca2+ channels are found (Budnik et
al., 2006). On the postsynaptic side, glutamate receptors (GluRIIA and GluRIIB
subunits) are clustered in opposition to these active zone and are critical for increasing
synaptic strength (Marrus et al., 2004). Presynaptic input is required for this clustering to
occur, evidenced by diffuse GluR distribution (absence of receptor clustering) when
action potential firing is chemically or genetically blocked (Broadie and Bate, 1993b;
Broadie and Bate, 1993c). Interestingly, clustering can occur even in mutants that do not
release glutamate, which implies that synaptic transmission, but not glutamatergic input
is required (Daniels et al., 2006).
7
After hatching, Drosophila progress through three larval stages (named first through third instar stages). During this time, muscle fiber volume increases 150-fold,
and the NMJ must grow to provide sufficient synaptic input to drive muscle contraction
(Gramates and Budnik, 1999). Homeostasis of synaptic strength is achieved by the
constant addition of synaptic boutons and the increase in the number of active zones that
is coordinated to muscle growth (Davis, 2006). The axon ending grows along the length
of the muscle, expanding the absolute length of the NMJ. Several branch points form at
distal boutons that normally give rise to two synaptic boutons, further increasing synaptic
complexity. In addition, a structure called the subsynaptic reticulum (SSR) develops
during later larval stages and circumscribes the postsynaptic membrane of boutons. The
SSR is formed by expansion of the postsynaptic membrane resulting in a complex,
involuted system of muscle membrane that surrounds the entire bouton (Budnik, 1996).
The NMJ continues to expand as muscles enlarge, and neural activity is required for
appropriate synaptic growth (Budnik et al., 1990). NMJ growth patterns adapt to muscle
demand, with increased outgrowth observed with greater locomotor activity (Sigrist et al.,
2003). Furthermore, synaptic activity is necessary for the refinement of NMJs since excess connections are formed in the absence of neural activity during late embryonic and first instar larval stages (Jarecki and Keshishian, 1995). Although not required for initial synapse formation, postsynaptic glutamate receptors are necessary for maturation
of the postsynaptic NMJ machinery, evidenced by disorganized ultrastructure of
postsynaptic densities in GluRIIA hypomorphs (Schmid et al., 2006). The growth and development of Drosophila NMJ are highly stereotyped, predictable and reproducible.
Importantly, the NMJ is structurally and functionally adaptable making it an amenable
8
model system for the effect of genetic and environment perturbations on stereotyped
developmental pattern and synaptic activity.
Regulation of Drosophila NMJ morphology
Transsynaptic signaling regulates synaptic growth:
Presynaptic (anterograde signaling)
The Drosophila Wnt, Wingless (Wg) ligand is an anterograde signal that is secreted from the presynaptic motorneuron and taken up by the postsynaptic muscle by endocytosis (Packard et al., 2002). The Wg signaling pathway promotes stabilization and growth of synaptic boutons (Packard et al., 2002). In the postsynaptic compartment, Wg binds to its receptor, Drosophila Frizzled-2 (DFz2), which induces internalization of the receptor and transport to the muscle perinuclear area. The PDZ domain containing protein, dGRIP, interacts with DFz2 and promotes intracellular trafficking to the nucleus
(Ataman et al., 2006). The DFz2 C-terminal domain is then cleaved off and imported into the nucleus where it presumably regulates transcription during synapse development
(Mathew et al., 2005). Loss-of-function mutations in DFz2 and dGRIP inhibit synaptic bouton formation and induce a "ghost" phenomenon where boutons lack identifiable active zones, postsynaptic densities, and SSR (Ataman et al., 2006; Mathew et al., 2005).
Thus, the Wg signaling pathway is necessary for the maintenance of normal synapses.
9
Postsynaptic (retrograde signaling)
The absence of appropriate muscle targets does not prohibit the migration of motorneurons to their normal target areas nor the formation of presynaptic active zone structures (Prokop et al., 1996). However, retrograde feedback signals from the muscle back to the neuron are important for maturation and regulation of structural plasticity of the synapse. Adaptive changes would necessitate that changes in muscle activity would be communicated to motorneurons. In this way, the quantity and quality of synaptic firing can be titrated.
The best characterized retrograde signal identified thus far is Glass Bottom Boat
(Gbb), a Drosophila member of the bone morphogenetic protein (BMP) family (McCabe et al., 2003). Gbb is released from the muscle and binds to the type I BMP/TGFβ receptors, Thick veins (Tkv) or Saxophone (Sax) or to the type II receptor, Wishful thinking (Wit), expressed on motorneurons (Aberle et al., 2002; Marques et al., 2002;
Rawson et al., 2003). Dimerization of type I and II receptors induces phosphorylation of
Mothers against decapentaplegic (MAD), a Receptor-activated Smad (R-Smad) transcription factor (Rawson et al., 2003). The co-Smad, Medea, then facilitates translocation of phospho-MAD back to the neuronal nucleus where it directs the transcription of structure modifying genes (Ball et al., 2010). One of these genes is the
Rho-type guanyl-nucleotide exchange factor, Trio. Together with Rac, Trio is thought to promote NMJ growth by regulating the actin cytoskeleton (Ball et al., 2010). Mutations in Gbb and the BMP receptors result in decreased bouton numbers and NMJ size (Aberle et al., 2002; McCabe et al., 2003). Furthermore, synaptic ultrastructure is abnormal and neurotransmitter release is reduced (Keshishian and Kim, 2004).
10
Whereas, Gbb promotes synaptic growth, the E3 ubiquitin ligase, Highwire
(HiW) may be necessary to restrain NMJ inappropriate expansion. hiw mutants have dramatically larger synaptic size and increased numbers of synaptic boutons (Wan et al.,
2000). HiW binds to the Smad protein Medea (Med) and may negatively regulate the
BMP signaling cascade. Mutations in the BMP pathway suppress the hiw loss-of-
function phenotype, and increased BMP signaling in hiw mutants induces synaptic
expansion (McCabe et al., 2004). Additionally, the F-box protein, DFsn, binds to HiW to
from a ubiquitin ligase complex. Together they down-regulate the activity of the
mitogen-activated protein (MAP) kinase kinase kinase (MAPKKK) Wallenda (Wu et al.,
2007). Wallenda is a Drosophila homologue to vertebrate dual leucine kinase and downstream signaling target of HiW (Collins et al., 2006). Mutational analysis shows that loss of DFsn phenocopies hiw mutation and that hiw and DFsn genetically interact
(Wu et al., 2007). These results suggest a balance between BMP signaling and regulation by Highwire/DFsn, governing the growth of the NMJ.
Actin cytoskeleton dependent plasticity
Presumably, the regulators of morphological plasticity must exert their actions on physical effectors to modulate synaptic growth. These molecular effectors will consequently "shape" the synapse. Elements of the cytoskeleton are obvious candidates to fulfill this role. Although both the microtubule and actin cytoskeletons have been identified as critical conversion points for synaptic remodeling in vertebrates and
Drosophila, I will focus solely on the actin cytoskeleton for the purposes of this review.
11
The actin cytoskeleton has been implicated to play key roles in the development
and plasticity of the brain. For example, spatial segregation of different synaptic vesicles
pools are maintained by scaffolding to actin (Cingolani and Goda, 2008). Additionally,
actin remodeling may allow for the physical trafficking of synaptic vesicles between
functional pools (Cingolani and Goda, 2008; Hotulainen and Hoogenraad, 2010). On the
postsynaptic membrane, neurotransmitter receptors are clustered by anchoring proteins
that associate with the postsynaptic density and actin cytoskeleton (Kuriu et al., 2006). In
Drosophila, the 4.1 protein, coracle, fulfills this role. Coracle binds to the C-terminus of the GluRIIA subunit (but not GluRIIB) and facilitates postsynaptic localization and anchoring of glutamate receptors (Chen et al., 2005). In loss of function coracle mutants, type A glutamate receptors are lost with no detectable change in type B receptor function or localization. Furthermore, pharmacologic disruption of postsynaptic actin in wild type
flies phenocopies loss of coracle by reducing GluRIIA (Chen et al., 2005). Because actin
is the most prominent cytoskeletal element in both pre- and the postsynaptic terminals,
modulation of actin dynamics is likely to drive morphologic and functional plasticity of
the synapse (Li et al., 2010).
Dynamic actin remodeling not only regulates synaptic efficacy and the
localization of synaptic machinery, but it also induces visible morphological changes.
Actin has prominent roles in neurite formation, extension and branching, as well as in
establishment of new synapses (Cingolani and Goda, 2008). Notably, the actin
cytoskeleton is thought to be critical for the regulation of dendritic spines. Dendritic
spines are small postsynaptic protrusions that mediate most of the excitatory synaptic
transmission in the brain (Matus, 2000; Yuste and Bonhoeffer, 2004). They are a major
12
site of information processing in the brain because long lasting structural changes, such
as enlargement of dendritic spines, form the basis for long-term potentiation (LTP)
(Bramham, 2008). Durable forms of actin-mediated synaptic plasticity have been proposed as mechanisms to encode memory. For example, induction of LTP induces the shift from monomeric (G-actin) to polymerized, or filamentous actin (F-actin) in dendritic spines. This results in increased in spine volume (Okamoto et al., 2004). The converse pattern is true for long-term depression (LTD), where spine volume is decreased.
Proteins that bind to and modulate actin dynamics have been proposed to regulate structural plasticity of the synapses. The Arp2/3 complex is a stable complex consisting of seven subunits including two actin related proteins Arp2 and Arp3 in complex with
ARPC1-5. The Arp2/3 complex is the primary microfilament-nucleating machinery in eukaryotic cells and is necessary for formation of lamellapodia (Hotulainen and
Hoogenraad, 2010; Pollard and Beltzner, 2002). In addition, branched F-actin networks are initiated by Arp2/3, which can bind to the sides of existing filaments and form nascent nucleation points (Goley and Welch, 2006). Arp2/3 is highly enriched in dendritic spines and knockdown of Arp2/3 impairs the formation of dendritic spine heads
(Racz and Weinberg, 2008). Furthermore, inhibition of Arp2/3 activators including Abi,
WAVE-1 and N-WASP result in decreased dendritic spine numbers and altered morphology (Grove et al., 2004; Kim et al., 2006; Wegner et al., 2008). Cytoskeletal growth is controlled by small Rho GTPases in a wide range of organisms (Hall, 1998).
N-WASP, or Neural Wiskott-Aldrich Syndrome protein, is one of these family members.
N-WASP binds and activates Arp2/3 by cooperating with Cdc42. By binding Arp2/3
13
complex via its C-terminus, N-WASP forms a functional link between signaling and actin polymerization (Rohatgi et al., 1999). Furthermore, N-WASP has been suggested to be able to transiently bridge actin filaments directly to lipid membrane via its WASP homology 2 (WH2) domain, adding another possible mechanism to change cell shape
(Co et al., 2007). WASP regulation of Arp2/3 is necessary for lamellapodia and filopodia formation and extension during axonal and dendritic growth in mammals (Banzai et al.,
2000). Hence, regulators of the actin cytoskeleton are clearly involved in morphologic plasticity of vertebrate neurons.
There are several lines of evidence implicating a role for actin in regulating synaptic structure. Spectrin is an F-actin crosslinker that was originally identified to be critical for the maintenance of erythrocyte shape (Bennett and Baines, 2001). In
Drosophila, presynaptic spectrin is necessary for synapse maintenance. Cell autonomous knockdown of spectrin in motorneurons leads to disassembly of the synapse (Pielage et al., 2005). This presumably occurs because spectrin stabilizes cell adhesion molecules such as neuroglian and fasciclin II (Pielage et al., 2005). Postsynaptically, spectrin forms a scaffold with actin to coordinate development. Loss of postsynaptic spectrin results in abnormally large and irregularly spaced active zones. Furthermore, spectrin mutations negatively impact growth by reducing the number of synaptic boutons and formation of the SSR (Pielage et al., 2006). The functional consequences of this defect is an increase in quantal size without changes in presynaptic vesicle size, perhaps as an attempt to compensate for the deficient numbers of synaptic boutons.
14
The function of Drosophila Wsp and linking actin to NMJ growth
The Drosophila homologues of Arp2/3 activators have also been identified and
have been shown to be involved in many actin-related functions. Like its vertebrate counterpart, Drosophila WASp (Wsp) acts as a potent activator of Arp2/3 in certain developmental context (Ben-Yaacov et al., 2001). It has been shown to be necessary for correct microvillus formation in rhabdomeric photoreceptor, myoblast fusion during
muscle development, and mediating cell fate decisions in sensory organs. In
photoreceptors, Wsp accumulates on the apical surface before microvillus formation,
which coincides with apical F-actin increase. Mutations in wsp lead to malformed
rhabdomeres caused by delayed F-actin accumulation and delayed formation of
primordial microvillar projections (Zelhof and Hardy, 2004). During myogenesis,
formation of syncytial muscle fibers occurs by repeated rounds of myoblast fusion. Wsp
interacts with the muscle specific protein D-WIP, a Drosophila homolog to
Verprolin/WASP interacting protein (Berger et al., 2008). D-WIP then allows
association with cell adhesion molecules necessary for the fusion events (Massarwa et al.,
2007). At the Drosophila NMJ, Wsp is expressed both pre- and postsynaptically and is
important for regulating synaptic growth (Coyle et al., 2004).
The discovery of the F-BAR and src-homology 3 (SH3) domain containing
protein, nervous wreck (nwk) provide a mechanistic link between BMP signaling, endocytosis, and synaptic growth via modulation of actin (Collins and DiAntonio, 2004;
O'Connor-Giles et al., 2008; Rodal et al., 2008). Nwk is expressed in motor neurons terminals (presynaptic) and localizes to periactive zones in synaptic boutons. Nwk loss-
of-function mutations induce abnormal NMJ development with expanded growth and
15
excessive arborization compared to wild type (Coyle et al., 2004). Furthermore, mutants
display a distinctive morphological phenotype of hyperbranched boutons. Normally, single NMJ branch points give rise to only two boutons, but nwk mutants, have boutons that give rise to three, four and even five boutons (Coyle et al., 2004). Nwk interacts with Wsp and activates Wsp-Arp2/3 mediated actin polymerization via its first (of two)
SH3 domains (Coyle et al., 2004; Rodal et al., 2008). Presynaptic transgenic expression of nwk rescues the loss-of-function phenotype. Finally, nwk and wsp genetically interact
in a dose dependent manner as evidenced phenotypic enhancement in flies mutant for
nwk and wsp (Coyle et al., 2004). Genes that do not directly influence each other rarely
display this type of genetic interaction. Recent studies have further implicated the
convergence of endocytosis and BMP signaling with actin assembly for the nwk based
regulation of synaptic growth. In addition to a genetic interaction with wsp, loss of nwk
is exacerbated by endocytic mutations (endo, dap160, and shi). Nwk interacts with
components of endocytic machinery, Dap160 and dynamin. Specifically, nwk binds the
SH3 domain of Dap160 and the proline-rich domain (PRD) of dynamin and could
potentially form a multiprotein complex with these proteins (O'Connor-Giles et al.,
2008). The nwk overgrowth phenotype is dependent on BMP signaling. Loss of the Gbb
(BMP ligand) receptor wit, results in undergrown NMJ and nwk, wit double mutants are
indistinguishable from wit only mutants (Rodal et al., 2008). Additionally, nwk interacts
with and negatively regulates another BMP receptor, tkv (O'Connor-Giles et al., 2008).
Finally, Nwk cooperates with the Rho GTPase, Cdc42, to promote Wsp-dependent actin
polymerization. Cdc42 has a punctate distribution both pre- and postsynaptically, where
it colocalizes with Nwk on the neuronal membrane (Rodal et al., 2008). The relationship
16
between actin cytoskeletal dynamics and endocytosis is not surprising since the Arp2/3 complex is known to generate kinetic forces at various steps in the endocytic pathway
(Kaksonen et al., 2006). Furthermore, endocytosis mutants display NMJ morphology defects, and this phenomenon may be related to changes in actin dynamics (described in more detail in the next section). However, it is unclear whether endocytosis is a morphogenic mechanism or if structural changes are compensatory adaptations to mutations altering synaptic transmission.
Endocytic proteins are negative regulators of NMJ growth
Several proteins associated with endocytosis and synaptic vesicle retrieval also play important roles in regulating synaptic architecture in Drosophila. Dynamin- associated protein 160 kDa (Dap106, also known as Intersectin) interacts with dynamin, as its name implies, and acts to scaffold endocytic machinery at synapses (Broadie,
2004). Dynamin I is a GTPase that catalyzes the fission step to form distinct endocytic vesicles from clatharin-coated pits (Sweitzer and Hinshaw, 1998). Loss of dap160 in
Drosophila causes a synaptic vesicle endocytosis (SVE) defect that is most pronounced at higher rearing temperatures (implying an increase in synaptic activity). In addition, the
NMJ of dap160 mutants display increased branching and numbers of synaptic boutons, many of which are small in size (Koh et al., 2004; Marie et al., 2004). Multiple
"satellite" boutons arise from branch points that should only give rise to two boutons
(hyperbranched boutons). NMJ span, however, relative to the muscle is unaffected in these mutants (Marie et al., 2004). Loss of other endocytosis genes show consistent changes in NMJ morphology. Mutations in dynamin (shibire), endophilin,
17
synaptotagmin, synaptojanin and AP180 all show increased synaptic growth with satellite boutons (Dickman et al., 2006). Increase in branching and density of synaptic boutons is also observed with the loss of rab11 (Khodosh et al., 2006). Rab11 plays important roles in the exocytic biosynthetic pathway as well as SVE. These NMJs resemble those observed in dap160 mutants. Epidermal growth factor receptor pathway substrate clone
15 (Eps15) is a protein implicated in endocytosis, endosomal trafficking and cytoskeletal regulation. NMJs of eps15 null flies show pronounced increase in bouton number, branch points and the abnormal presence of hyperbranched boutons (Koh et al., 2007).
Spinster (spin), a multipass transmembrane domain and late endosomal protein, regulates synaptic growth and presynaptic transmitter release (Sweeney and Davis, 2002). Spin mutants have twice the normal number synaptic boutons at their NMJs and additionally have deficits in presynaptic release. Interestingly, additional mutations to components of the BMP pathway (thv, sax and wit), suppress the spin phenotype and in some double mutants, the BMP receptor undergrowth phenotype is dominant (Sweeney and Davis,
2002).
The etiology of satellite bouton formation as it relates to endocytosis is not clearly understood. Ultrastructural synaptic changes are observed in the synapses of endocytosis mutants. Loss of dap160 causes a depletion of endocytic proteins in addition to decreasing the number and changing the size of synaptic vesicles (Koh et al., 2004).
However, satellite boutons that appear in endocytosis mutants do appear to be functional, since they contain releasable vesicles and a normal complement of synaptic proteins.
Synaptojanin mutants contain more active zones per membrane area, perhaps as a compensatory mechanism for inefficient transmission (Dickman et al., 2006). Whereas
18
Spin mutants show expansion of the late endosomal/lysosomal compartment (Sweeney
and Davis, 2002). One possibility is that expanded NMJ with supernumerary boutons
form as an adaptation to altered synaptic transmission, since all endocytosis mutations
alter the release of neurotransmitter. However, satellite boutons appear to form
independently of synaptic transmission and glutamatergic signal. Transmission blockade
by transgenic expression of tetanus toxin is insufficient to induce supernumerary bouton
formation or block them from forming in endocytic mutants (Dickman et al., 2006).
Furthermore, the anatomical phenotype is not observed in glutamate receptor mutants
(Dickman et al., 2006). It is of note that the structural phenotype of endocytosis mutants
strongly resembles that of nwk (Coyle et al., 2004). Nwk binds to many of these
endocytosis proteins and may at least in part act by negatively regulating the BMP pathway (O'Connor-Giles et al., 2008). This raises the possibility that endocytosis influences synaptic growth not necessarily by affecting functional transmission, but by regulating the insertion or removal of signaling molecules from the plasma membrane,
like components of the BMP pathway. Direct evidence to support this hypothesis has not
been shown, but phenocopy of nwk and endocytosis mutants and suppression of spin
phenotype by additional mutations in the BMP cascade support this assertion. Another possibility is simply that endocytic proteins may have distinct functions depending in
which protein complex they are found. Many proteins involved in endocytosis contain an
N-terminal F-BAR domain that can bind and induce tubulation of lipid membranes
(discussed later) in addition to domains mediating protein-protein interactions. Hence, it may be possible for a single "endocytosis protein" to scaffold dynamin to the cytoskeleton (via interaction with N-WASP) during endocytosis; bind synaptic vesicle
19
and N-WASP to localize and physically traffic vesicles intracellularly; and link the actin cytoskeleton to plasma membrane, perhaps inducing morphological changes.
Vertebrate Syndapin
Synaptic dynamin associated proteins, or Syndapins (also known as PACSINs in
Human or FAP52 in chicken), are highly conserved proteins that are expressed in all multicellular eukaryotes (Kessels and Qualmann, 2004). Syndapins contain an N- terminal FCH-BIN amphiphysin RVS domain (F-BAR), a central coiled-coil domain, three NPF repeats and C-terminal SH3 domain (Halbach et al., 2007; Qualmann et al.,
1999). Syndapins have been proposed to link membrane trafficking with the cytoskeleton by being able to interact with both endocytic proteins and regulators of the actin cytoskeleton. In mammals, Syndapin exists in three isoforms: Syndapin I (Synd I) expression is restricted to the brain, Synd II is ubiquitously expressed and Synd III is found predominantly in skeletal and cardiac muscle (Modregger et al., 2000; Plomann et al., 1998; Ritter et al., 1999). All three isoforms binds to the large GTPase dynamin I and synaptojanin (Modregger et al., 2000). Furthermore, because Synd binds N-WASP via its SH3 domain, it presumably couples bursts of actin polymerization with endocytosis, providing a mechanism to physically move vesicles away from the cell membrane
(Qualmann and Kelly, 2000; Qualmann et al., 2000). Consistent with a role in actin dynamics, Syndapins localize to lamellapodia and filopodia, sites of high actin turnover
(Kessels and Qualmann, 2002). Unlike Nwk, Synd contains a single SH3 domain, which would presumably preclude Synd's ability to function as a scaffold, since concurrent binding of several different proteins would be required for this function. To overcome
20
this, Synd is able to self associate via its N-terminal FCH domain to form dimers or tetramers (Halbach et al., 2007; Kessels and Qualmann, 2006; Rao et al., 2010). Mutants that are unable to oligomerize are defective in their ability to mediate cytoskeletal rearrangements and endocytosis (Kessels and Qualmann, 2006).
Indeed, the function of Synd has been clearly demonstrated for synaptic vesicle endocytosis (SVE). Synaptic transmission involves the exocytosis of neurotransmitter filled vesicles. However, in order to maintain functional synapses over time, cell membrane has to be recycled by endocytosis to form new vesicles. This is critical to maintain correct numbers of synaptic vesicles and also to maintain cell volume after neurotransmitter release. During endocytosis, the cell membrane invaginates with the help of proteins like clatharin. The enzymatic activity of dynamin I is then required for the fission step to form an endocytic vesicle (Sweitzer and Hinshaw, 1998). Dynamin's interactions with Synd I, amphiphysin I and endophilin I have been implicated in SVE.
Dynamin is normally dephosphorylated during SVE, but in synaptosomes, dynamin is phosphorylated at serine 774 and 778 and this induces calcineurin-dependent interaction with Synd (Anggono et al., 2006). Although both Synd and endophilin bind the same region of dynamin, phosphorylation states determines preferential interaction of one over the other (Anggono and Robinson, 2007). Inhibition of Synd disrupts SVE under conditions of high frequency stimulation (5 Hz) (Andersson et al., 2008). Interestingly, under low frequency stimulation (0.2 Hz), disruption of Synd does not affect SVE in contrast to what is observed with amphiphysin inhibition (Andersson et al., 2008;
Shupliakov et al., 1997). Perhaps, this suggests divergent non-redundant functional roles for the dynamin and N-WASP interacting proteins, Synd and amphiphysin.
21
F-BAR Domain containing proteins: structural bridges between cytoskeleton and
cell membranes
In addition to its ability to bind to endocytic proteins, Synd has been demonstrated
to associate with lipid membrane directly via its F-BAR (FCH [Fer/Cip4 homology]-Bin-
Amphiphysin-Rvs) domain. F-BAR containing proteins are cytosolic proteins with intrinsic membrane-deforming and stabilizing properties (Itoh et al., 2005). F-BAR domains themselves can tubulate membrane in vitro and have been suggested to be involved in diverse cellular functions such as T-tubule morphogenesis, vesicle endocytosis, and cell migration (Dawson et al., 2006; Frost et al., 2009; Habermann,
2004; Shimada et al., 2007). The F-BAR domain of Synd efficiently deforms liposomes into tubules in vitro (Rao et al., 2010; Wang et al., 2009). However, recombinant full- length synd is unable to induce tubulation. Rather, a truncation mutant lacking the SH3 domain or the F-BAR domain alone are able to induce tubulation. Interestingly, the F-
BAR and SH3 domains interact and this interaction may autoinhibit the membrane
deformation function of Synd (Rao et al., 2010). Membrane tubulation can be stimulated
when full length Synd is incubated with a dynamin peptide, which presumably binds to
the SH3 domain and displaces F-BAR. Furthermore, full length synd containing a SH3
mutation abrogating SH3 to F-BAR binding is able to tubulate liposomes by itself (Rao et
al., 2010). Overexpression of the F-BAR domain of Synd II in HeLa cells induces
formation of intracellular tubules and cellular microspikes, which are membrane
protrusions (Shimada et al., 2010). Microspike generation is dependent on actin
polymerization and Synd II F-BAR is concentrated in the neck of the microspikes
(Shimada et al., 2010). Taken together, these data suggest that Synd has the functional
22
potential to bind and shape membrane by interacting with actin nucleation machinery and directly binding membrane. Interaction with SH3 domain binding proteins may be necessary to relieve autoinhibition and function in this capacity. How the membrane shaping function directly relates to its ability to bind to endocytic protein and N-WASP is still unclear, but the discovery of F-BAR function in Synd bolsters its role a cellular adaptor protein.
An important functional consequence of F-BAR mediated membrane association is the modulation of neuromorphogenesis. Expression of Synd in cultured hippocampal neurons induces dendrite formation and branching (Dharmalingam et al., 2009). The
SH3 domain is necessary, but not sufficient to induce these morphological changes.
Furthermore, Synd induces N-WASP mediated actin polymerization, which is enhanced by Synd targeting to the plasma membrane. The SH3 domain mediates cortical actin rearrangement, but only if expressed as full length Synd, tagged with the F-BAR, or tagged with other plasma membrane targeting domains (Dharmalingam et al., 2009).
Surprisingly, morphological effects of Synd expression do not appear to be dependent on changes to endocytosis or endosome retrieval. Treatments known to block endocytosis fail to phenocopy the overexpression of Synd in hippocampal neurons (Dharmalingam et al., 2009). Finally, RNAi knockdown of Synd results in increased axon length and branching in hippocampal neurons. Knockdown of N-WASP, Arp2/3 and Abp1 phenocopy the loss of Synd strongly suggesting that regulation of neuron morphology is dependent on interaction with N-WASP and actin polymerization .
Synd also plays a role in development in vertebrates. The Zebrafish Synd ortholog, Pacsin 3, has been shown to be necessary for formation of the notochord, the
23
defining feature of all chordates (Edeling et al., 2009). During development, the
notochord releases the morphogen Sonic Hedgehog (Shh), which is critical for
establishing polarity of the neural tube). Pacsin 3 clearly localizes in the notochord in
Zebrafish. Injection of pacsin3 antisense morpholino oligonucleotide (MO; resulting in gene knockdown) results in severe developmental abnormalities that are dose-dependent.
Ectopic injection of Drosophila synd RNA rescues the pacsin3 MO phenotype.
However, mutant Drosophila synd RNAs in which the SH3 domain is truncated or
contain F-BAR mutations that abrogate membrane association fail to rescue the loss of
pacsin3 (Edeling et al., 2009). This demonstrates that the SH3 domain and membrane
association are necessary for a developmental function of vertebrate synd.
Drosophila Syndapin
Drosophila have a single Syndapin gene (synd). Drosophila synd is expressed widely in actin rich structures and at the NMJ (our unpublished data). The Drosophila synd SH3 domain binds to the Drosophila ortholog of N-WASP (wsp) and also the proline rich domain (PRD) of dynamin (our unpublished data) (Kumar et al., 2009a).
Additionally, the crystal structure of the Drosophila synd F-BAR domain shows structural resemblance with vertebrate synd (Edeling et al., 2009). Synd F-BAR domains assemble as dimers and form a structure resembling an elongated bowl. Hydrophobic
(membrane associating) residues line the concave aspect of this "bowl" (Edeling et al.,
2009). In S2 cultured cells, expression of both the F-BAR domain and full length Synd cause intracellular tubulation, though the full length protein induces much weaker tubulation (Kumar et al., 2009b). Membrane binding deficient mutants or SH3 alone fail
24
to induce tubule formation (Kumar et al., 2009b). These characteristics suggest analogous roles for Drosophila synd when compared to its vertebrate orthologs.
In contrast to vertebrate Synd, at the Drosophila NMJ, Synd is predominantly expressed postsynaptically, weakening its theoretical potential for involvement in synaptic vesicle recycling (our unpublished data) (Kumar et al., 2009a). It is important to note that based on the resolution of confocal microscopy, it cannot be ruled out that Synd localizes in motorneurons very close to the membrane since the pre- and postsynaptic membranes are in such close apposition. Synd colocalizes with Wsp, which is present presynaptically, but is predominantly expressed in the muscle (Coyle et al., 2004; Kumar et al., 2009b). Surprisingly, analysis of a synd hypomorphic line does not reveal defect in either synaptic transmission or SVE (Kumar et al., 2009a). Rather, a morphological phenotype is uncovered when it is overexpressed postsynaptically. The postsynaptic SSR is greatly expanded when synd is misexpressed (Kumar et al., 2009b). The SH3 domain is necessary for targeting to the synaptic membrane and both SH3 and F-BAR are necessary to induce NMJ SSR expansion. Interestingly, membrane expansion can be induced even in Dlg and dPAK mutants, two known regulators of SSR development, suggesting that overexpressed synd acts independently or downstream of these known regulators (Kumar et al., 2009b).
25
Research aims
The main goal of Part I of this dissertation is the analysis of the function of
Drosophila Syndapin (Synd) at the neuromuscular junction. We postulated that Synd regulates cytoskeletal dynamics at the synapse by interacting with Wiskott-Aldrich
Syndrome protein (Wsp). Thus, we wanted to first show that like mammalian forms of
Synd, the SH3 domain of Drosophila Synd binds to the proline rich domain (PRD) of
Wsp. Then, we needed to determine where Synd localizes at the synapse and determine if Synd is expressed pre- or postsynaptically. Also its proximity to the synaptic membrane and relative distribution with respect to Wsp needed to be determined. In order to analyze the Synd function, we aimed to generate loss-of-function mutations by imprecise excision of transposon insertion lines. Since mammalian Synd has been suggested regulate neuronal morphology, we wanted to analyze the Drosophila synd mutants for defects in stereotyped synaptic architecture. Definitive implication of synd in generation of structural phenotype would be determined by transgenic rescue of the loss- of-function phenotypes. The cell types of synd action could then be determined by driving rescue constructs using specific promoters. Since we suspect the biochemical association between Synd and Wsp to be important for regulating synaptic plasticity, we wanted to go further to determine if synd and wsp genetically interact by making double mutant flies. Finally, because Synd may function as an adaptor protein by recruiting
Wsp, we sought to examine the NMJ distribution of Wsp in a synd mutant, and conversely analyze Synd localization in a wsp mutant. This would suggest a functional relationship between the two genes, but also reveal epistasis of gene interaction. We hypothesize that that in Drosophila, Synd recruits Wsp to the postsynaptic membrane of
26
the neuromuscular junction, and that this recruitment is important to regulate synaptic development.
27
CHAPTER 2
DROSOPHILA SYNDAPIN INTERACTS WITH WISKOTT-
ALRICH SYNDROME PROTEIN AND REGULATES
NEUROMUSCULAR JUNCTION GROWTH
POSTSYNAPTICALLY
28
Introduction
Synaptic plasticity describes the ability of neurons to undergo adaptive modifications and is believed to underlie complex processes such as long-term potentiation and memory storage. Synapses are capable of rapid structural reorganization in response to stimuli (Fischer et al., 2000; Zito and Svoboda, 2002). Hence, morphologic changes have been suggested to contribute to neuronal plasticity, even in the short-term. During glutamatergic neuromuscular junction (NMJ) development in
Drosophila, the surface area of the postsynaptic muscle increases over 100 fold (Prokop and Meinertzhagen, 2006). The motorneuron consequently increases the number of active zone containing synaptic boutons to compensate for the increased demand. This process somewhat resembles yeast budding where new boutons arise from pre-existing boutons (Zito et al., 1999). Synapses are further modified in an activity-dependent manner by structural and functional changes (Broadie and Bate, 1993a; Budnik et al.,
1990; Sigrist et al., 2003). Synaptic growth is highly regulated resulting in a very stereotyped morphologic pattern defined by nerve entry point, branching pattern, and terminal size (Broadie and Bate, 1993b; Johansen et al., 1989). Considering the ease genetic manipulation of Drosophila and accessibility of these synapses, the NMJ is an ideal model to study modulators of synaptic development and plasticity.
The importance of the actin cytoskeleton has been well described for growth cone guidance and axon outgrowth presynaptically and dendritic spine formation postsynaptically (Banzai et al., 2000; Bogdan et al., 2004; Hotulainen and Hoogenraad,
2010; Racz and Weinberg, 2008). Recently, evidence for its role in the regulation of synaptic growth has been growing (Wegner et al., 2008). Since both pre- and
29
postsynaptic compartments are enriched in F-actin, regulators of synapse architecture could presumably act by altering actin dynamics (Li et al., 2010). One such actin regulator, a presynaptic adaptor called nervous wreck (nwk), has been identified (Coyle et al., 2004). Nwk mutants display an overgrowth of NMJ with excess branching. Nwk also binds the Drosophila homologue of Wiskott-Aldrich Syndrome protein (Wsp), (Coyle et al., 2004). Wsp and its homologs, promote actin polymerization at the leading edge of motile cells by activating the actin-related protein (ARP2/3) complex (Higgs and Pollard,
2001; Rohatgi et al., 1999). This regulation has been demonstrated to be essential for the formation of lamellipodia and filopodia during neuronal growth in mammals. Indeed,
Nwk functionally cooperates with Cdc42 to promote Wsp-mediated actin polymerization.
Hence, nwk was proposed to form a signaling complex with Wsp and regulate synaptic growth. However, Wsp is expressed both pre and postsynaptically, and proteins that promote postsynaptic Wsp localization have yet to be identified. We believe that the
Drosophila homologue of Syndapin (Synd) fulfills this role.
Mammalian Syndapin I (Synd I, also called PACSIN and FAP52) is found in presynaptic nerve terminals and was shown to interact with the proline rich domains
(PRD) of neural-Wiskott-Aldrich Syndrome protein (N-WASP) and the endocytic protein, dynamin. Thus, Synd I potentially links synaptic vesicle endocytosis (SVE) with the actin cytoskeleton (Qualmann et al., 1999). The C-terminal src homology 3 (SH3) domain of Syndapin I mediates binding to dynamin, N-WASP and other synaptic proteins
(Qualmann et al., 1999). Several SH3 domains containing proteins that bind dynamin have been proposed to link signaling with molecules associated with the cytoskeleton
(Kessels and Qualmann, 2004).
30
Here we have identified the Drosophila ortholog of Synd, which is expressed
throughout actin rich structures and at the NMJ. Like its vertebrate counterparts, Synd
SH3 domain binds to the proline rich domain (PRD) of Drosophila Wiskott-Aldrich
Syndrome protein (Wsp). At the NMJ, Synd is expressed predominantly
postsynaptically, which was surprising considering its putative role in synaptic vesicle
endocytosis. In order to analyze synd function, we generated loss-of-function mutations
by imprecise excision of transposon insertion lines. Since mammalian Synd has been
suggested to modulate neuronal morphology, we wanted to demonstrate a role for
Drosophila synd in regulating structural plasticity. Synd mutants have overgrown NMJs
that resemble loss of nwk, wsp and several other genes related to endocytosis. The
structural phenotype is specific to postsynaptic synd since transgenic rescue of the loss-
of-function phenotypes was achieved by driving recombinant synd in the muscle
(postsynapse). Given the biochemical association between Synd and Wsp, we next show that synd and wsp genetically interact in a dose dependent manner by making doubly
mutant flies. Finally, because Synd may function as an adaptor protein by recruiting
Wsp, we examined the distribution and levels of Wsp in a synd mutant and conversely
Synd localization in a wsp mutant. Postsynaptic Wsp was lost in the synd mutant whereas no change in Synd was observed in a wsp mutant. These findings suggest that synd acts upstream of wsp and may function to bind and recruit Wsp. We hypothesize that this recruitment is important to regulate synaptic development.
31
Materials and Methods
Fly Stocks
Flies were maintained at 25º C in standard molasses media. EP(3)3506,
EP(3)0877, and EP(3)0409 were obtained from the Szeged Drosophila Stock Centre.
f07592 and c03868 were obtained from the Exelixis collection at Harvard. SyndDG10804
and Df(3R)BSC43 were obtained from the Bloomington Drosophila Stock Center. The
imprecise P-element excision alleles were generated by crossing EP(3)0877 with Dr,
transposase and screening potential null alleles by complementation tests with
EP(3)0409. Alleles with increased lethality were isolated and molecularly characterized
by PCR and sequencing. Precise excision lines were generated by isolating
chromosomes giving progeny in Mendelian ratio. A precise excision line, Precise105,
was confirmed by PCR, sequencing, and reversion of all tested phenotypes. This allele
was used as wild type control for phenotypic analysis. For the rescue experiments, the
muscle-specific driver MHC82-Gal4 and neural specific drivers Elav3A4-Gal4 and
Elav3E1-Gal4 were recombined with the B86.1F Syndapin allele. UAS-Synd was
generated by cloning a full length Synd cDNA (LD46328) into the EcoRI and StuI sites of pCaSpeR-hs. The Synd cDNA was cut from pCaSpeR-hs with EcoRI and SalI and cloned in to the EcoRI and XhoI sites of pUAST to generate UAS-Synd (Brand and
Perrimon, 1993). After germline transformation into yw67 embryos, several transgenic
lines were established. A UAS-Synd line on the X chromosome was crossed into the
Df(3R)BSC43 background for rescue.
32
Molecular Cloning and blot overlay assays
The mutant form of synd was generated by a fusion PCR strategy. The first fragment was amplified from pOT2-Synd (LD46328) using forward primer 5’-
ATGCAAATCCATTCGACGAG-3’ and reverse primer 5’-
CATAGTTGGCCAGATACAGTC-3’ and a second fragment was amplified using forward primer 5’-GGACTGTATCTGGCCAACTATG-3’ and reverse primer 5’-
ATCTGCATGTTCGTATCGGC-3’. These fragments were fused in a PCR reaction using the flanking primers, and the fusion product was cloned into the SacI and NdeI sites of pOT2-Synd to generate pOT2-SyndP487L. The SH3 domains of wild type and mutant
Synd were amplified using forward primer 5’-
CCGGAATTCTCTACCCCCAGACAAACAGC-3’ and reverse primer 5’-
GGCCAAGCTTACGCGGTCTCCACATA-3’ and cloned into EcoRI and HindIII sites of pGEX-KG to yield pGEX-SyndSH3 and pGEX-SyndSH3-P487L. GST fusion proteins were expressed in E. coli BL21(DE3) cells and purified using glutathione sepharose 4B beads (Amersham Biosciences) according to manufacturers protocol.
Amino acids 96-526 (encompassing the PRDs) of Drosophila Wsp were amplified from a
Wsp cDNA (Ben-Yaacov et al., 2001) using the forward primer 5’-
CCGGCATATGATCTGGGAGCACGAGATCTAC-3’ and reverse primer 5’-
GGTTGGATCCTTACTTACCACTCCCCTTCGTTGTC-3’ and then cloned into the
BamHI and NdeI sites of pET15b vector (Novagen). Recombinant Wsp was expressed in
BL21(DE3) cells in the presence (induced) or absence (uninduced) of 0.02 µM IPTG.
Equal amounts of bacterial protein extracts were subjected to SDS-PAGE and transferred to Immobilon-P membrane (Millipore). Blots containing Wsp were incubated with 0.2
33
µM GST or GST fusion protein and then probed with anti-GST antibody (Amersham
Biosciences, 1:1000) and then detected with HRP conjugated anti-goat antibody
(1:10000).
Immunohistochemistry and quantification of NMJ morphologic parameters
Since growth conditions can have significant effects on NMJ growth, crosses for a single experiment were set up in parallel and larvae were cultured on media from the same batch. Larvae of the correct genotype were selected and allowed to mature at a density of 50 per vial. Wandering 3rd instar larvae were dissected, fixed in 4%
paraformaldehyde (Sigma) in PBS for 30 min, permeabilized with PBST (PBS + 0.1%
triton X-100) for 10 min, and blocked overnight in 1% BSA in PBST at 4ºC. Larvae
were incubated in primary antibody diluted in blocking solution overnight, washed, and
incubated in Alexa-conjugated secondary antibodies diluted in blocking solution
overnight. Larvae were mounted in Vectashield (Vector laboratories) after extensive washing in blocking solution and PBST. Primary antibody dilutions are as follows: anti-
Synd, 1:5000; anti-syt (Mackler et al., 2002), 1:1000; anti-DLG (4F3; Developmental
Studies Hybridoma Bank), 1:500; anti-Wsp (Ben-Yaacov et al., 2001), 1:500; anti-CSP
and anti-HRP (Jackson ImmunoResearch Laboratories Inc), 1:200. Alexa 488, Alexa 555
or Alexa 633 conjugated secondary antibodies (Invitrogen) were used at 1:250.
Phalloidin conjugated to Alexa 488 (Invitrogen) was used at 1:200. All images were
captured on a confocal microscope (Zeiss LSM 510 META laser scanning module with a
Zeiss Axiovert 200M inverted microscope) with a 10x/0.30, 40x/1.20W, or 63x/1.40 oil
34
DIC objective. Confocal Z-sections were processed using Zeiss LSM 5 software (Rel.
3.2) and Adobe Photoshop CS2.
Quantification of NMJ morphology parameters was performed as previously described (Coyle et al., 2004). Briefly, larval NMJ 6/7 in segment A3 were stained with anti-syt and anti-HRP and photographed at 40x magnification. Synaptic boutons marked by distinct syt and HRP staining were counted by analyzing individual z-sections.
Branch points initiating from boutons downstream of the muscle entry point were scored.
Hyperbranched boutons were defined as boutons giving rise to 3 or more synaptic boutons (a bouton connected to 4 or more boutons). All values were normalized to the area of corresponding muscles 6 and 7, which was measured using phalloidin staining at
10x magnification. Synaptic lengths and muscle areas were computed using LSM 5 software. Statistical analysis was performed using Students' T-test and one-way ANOVA
(Tukey's post hoc test) in Prism 3.02 (GraphPad Software)
35
Results
Identification of a Drosophila Syndapin homologue.
We have used previously identified sequences of Syndapin (Qualmann and Kelly,
2000; Qualmann et al., 1999) and its homologs, PACSIN (Modregger et al., 2000;
Plomann et al., 1998; Ritter et al., 1999) and FAP52 (Merilainen et al., 1997), to search
databases for Drosophila synd homologs. A match was found corresponding to a single
curated gene, CG15693 (Lloyd et al., 2000) (Fig. 2A). Since the genome has been nearly
sequenced in its entirety, we can conclude that CG15693 represents the only synd gene in
the fly. The situation is more complicated in rats where there are two genes encoding
synd (Lloyd et al., 2000; Qualmann and Kelly, 2000; Qualmann et al., 1999), and in mice
where there are three genes encoding the protein (Modregger et al., 2000; Plomann et al.,
1998; Ritter et al., 1999). We chose to make loss of function alleles in Drosophila since
there is only one Syndapin gene present in the genome and we reasoned that reducing
expression of this single gene would likely result in a strong phenotype. Deleting only
one of the three mouse genes might not give a strong phenotype due to redundancy.
We sequenced a synd cDNA clone, LD46328, which contains a start codon, the entire open reading frame, a 5’ untranslated region and a polyA tail and hence is a full length clone. We have used the translation of this clone to determine that Drosophila
Syndapin (Synd) is a 494 amino acid protein that contains an N-terminal FCH
(FER/CIP4-homology) domain, coiled-coil, a NPF motif, and a C-terminal SH3 (src homology 3) domain (Figs. 1 and 2B) (Qualmann et al., 1999). The FCH domain has recently been re-categorized by including the coiled-coil loops and has been redubbed the
F-BAR domain (Itoh et al., 2005). Sequencing the full length Syndapin cDNA allowed
36
us to determine the molecular structure of the gene (Fig. 2A). It is made up of ten exons,
nine of which cover the open reading frame of the protein, and spans ~7kb of genomic
DNA. Alignment with mouse, rat, Xenopus and C. elegans forms of the protein show
that the fly isoform is 45% identical and 60% similar to the mouse isoform and is thus
well conserved. The protein is most highly conserved in its known functional domains
(see Fig. 1).
Identification P-element inserts near Drosophila Syndapin and characterization of
Syndapin hypomorphic alleles
We have identified five P-element insertions, within or close to the synd locus.
EP(3)0877, EP(3)0409, and DG10804 are inserted 2413 bp, 2044 bps, and 1903 bp
upstream of the start codon, respectively. f07592 and c03868 are inserted in the 3’ end of
the gene, 81bps and 462 bps from the stop codon (Fig. 2A). Locations relative to the start
codon are noted by the one-offset convention. P-element insertion points were confirmed
by PCR and sequencing. The EP(3)0877 insert is homozygous viable, but hatching of the
homozygotes is delayed compared to heterozygous siblings, suggesting that the insertion
in to the 5' exon has a detrimental effect on Synd expression. EP(3)0409, inserted in the
first intron, is essentially lethal; when larval density is kept very low some adult escapers
are seen. f07592 and c03868 are lethal but fully complement other synd mutants and the
chromosomal deficiency Df(3R)BSC43 (which is a synd null allele), suggesting that the
lethality may be due to other mutation(s) on the chromosome. The deficiency,
Df(3R)BSC43 (deletion of chromosomal segment 92F7-93B6) fails to complement synd
mutants, thus it completely removes the synd gene.
37
We have identified three Drosophila hypomorphic mutations caused by
transposon (P-element) insertions upstream of the synd start codon: EP(3)0877 in the first
non-coding exon, and EP(3)0409 and DG10804 in the first intron (Fig. 2A).
Homozygotes from these lines have reduced viability with few reaching adulthood.
Insertion points of P-elements inserted in, or near, the synd gene were confirmed by PCR
and sequencing (data not shown). To generate additional, potentially null alleles, the
EP(3)0877 line was remobilized by crossing it to a transposase source. We have
mobilized the EP(3)0877 P-element inserted at the 5’ end of the Syndapin gene to
generate a new Syndapin hypomorph, B86.1F (Fig. 2A). Breakpoints in these alleles have been determined by PCR and sequencing, which resulted in a 532 bp deletion to the right of the EP(3)0877 insertion site (2402 bp to 1870 bp upstream of the start codon), removing the entire first exon. B86.1F is lethal as a homozygote, with few adult
escapers. Adults that are B86.1F homozygous or B86.1F/Df are infertile, flightless and
display muscle tremors, which were not observed for EP(3)0877 homozygotes. Precise
excisions of the EP(3)0877 P-element are viable, have normal levels of Syndapin expression and display no obvious phenotypes (data not shown). Sequencing confirmed precise excision of the line, prec105, as well as precise excision revertants of EP(3)0409.
This reversion data suggests that phenotypes observed from heterozygotes of B86.1F and the deficiency line are unlikely to be due to second site mutations since they are derived from independent chromosomes. (Fig. 2A).
We raised an antibody against the C-terminal domain of fly Synd. The anti-Synd antibody appears to be specific as it recognizes a single ~66 kDa band. Preimmune serum does not cross react (data not shown) and together with the reduction in the
38
intensity of the protein band in the synd mutants suggests that the antibody raised is
specific for Synd. Western blots using fly head extracts from B86.1F as well as the other
mutant lines show decreased Synd relative to wild type (Fig. 2C and data not shown).
Furthermore, synd rescue lines (discussed later), in which recombinant synd is
transgenically expressed in mutant backgrounds show restoration of synd protein level as
assessed by the Synd antibody. B86.1F/Df is a synd hypomorph since adult flies contain
~1.8% of wild-type Synd protein levels (Fig. 2D).
Localization of Syndapin at Drosophila neuromuscular junctions.
Initial characterization showed that Drosophila Syndapin (Synd) is distributed
throughout several tissues in embryos and larvae. Synd is found at the morphogenic
cleavage furrow during cellularization in early embryogenesis; in the neuropil of the
embryonic nervous system (Fig. 3A); at the rhabdomere during eye development, with staining persisting into adulthood (Fig. 3B-D); and larval NMJ (Fig. 5). The expression pattern suggests that it may be involved in the processes of cellularization during embryonic development, eye development, and neuromuscular junction formation and/or maintenance. Synd is found in actin-rich tissues with highly curved membranes, such as the rhabdomere, thus suggesting a role as a regulator of cellular morphology.
Immunohistochemical analysis of Drosophila 3rd instar larvae was performed to
show Synd localization (Fig. 5A-C). Synd colocalizes with the postsynaptic marker, discs large (dlg), the Drosophila homolog to PSD-95, but only shows partial colocalization with a presynaptic marker, cysteine string protein (csp). Synd staining is observed to extend deeper into the muscle than even dlg. This indicates that Synd is
39
predominantly expressed postsynaptically at the NMJ, which is surprising for a protein with a hypothesized role in synaptic vesicle endocytosis. It should be noted that some dlg protein is found presynaptically (Lahey et al., 1994). This data clearly shows that Synd is found postsynaptically at the NMJ but using immunocytochemistry. However, at the level of light microscopy, we cannot rule out that Synd has a presynaptic distribution since the synapse gap is ~50 nm and the resolution of light microscopy is greater than
200 nm. Muscles are innervated by type-I synaptic boutons, which can be further subdivided into type-Ib (big) and type-Is (small) boutons (Gramates and Budnik, 1999).
Based on similar immunostaining to dlg, we conclude that Synd is present in all type-I boutons of NMJ 6/7 (Fig. 5G).
Loss of Syndapin results in synaptic overgrowth
In order to determine the in vivo synaptic function of Drosophila Synd, we analyzed loss of function mutants. We have already showed that mutants have decreased
Synd level in adult fly heads (Fig. 2D). Using our anti-Synd antibody we now show that
B86.1F/Df shows decreased levels of Synd at the larval NMJ compared to wild type controls (Fig. 5H). Likewise, the P-element mutant alleles EP(3)0877, EP(3)0409, and
DG10804 show decreased total Synd protein level and staining at the neuromuscular junction (data not shown). NMJ morphometric analysis was performed by co-staining 3rd instar larvae with anti-synaptotagmin (syt), a synaptic vesicle marker, and anti-Horse radish peroxidase (HRP), which labels the neuronal membrane in Drosophila.
Examination of B86.1F/Df revealed a striking synaptic overgrowth phenotype when compared to the precise excision control (prec105) (Fig. 6A & 6B). Since this phenotype
40
was consistent among several different independently derived allele combinations both as
homozygotes, transheterozygotes and hemizygotes with the deficiency, we deduced that
Synd loss was causing the phenotypes and not second site mutations that may be present
on the same chromosome (data not shown).
During development of glutamatergic synapses of type-I terminals, NMJs adopt a
stereotypic and reproducible NMJ growth pattern (Prokop, 2006). Morphometric
analysis was therefore performed on synd mutants. Specifically, we examined the
synaptic length, number of synaptic boutons, and number and complexity of branches of
NMJ 6/7 of segment A4. All values were normalized to area of the corresponding target
muscle since the NMJ grows proportionally to muscle size (Broadie and Bate, 1993d;
Schuster et al., 1996; Sigrist et al., 2003). NMJ morphological parameters of prec105
(the most appropriate wild type control for the imprecise excision) did not differ from
other wild type lines (w1118), indicating that precise P-element excision controls were
representative of a true wild type by reverting the synaptic growth phenotype. The span of NMJ 6/7 relative to muscle area was increased by approximately 60% in synd mutants
when compared to wild type (NMJ length per muscle area: prec105 = 1.91 ± 0.10, n = 18
versus B86.1F/Df = 3.03 ± 0.13 µm-1 x 103, n = 12; p < 0.001; Fig. 6D). In addition, the
number of synaptic boutons was increased by ~57% (prec105 = 1.08 ± 0.07, n = 18
versus B86.1F/Df = 1.70 ± 0.14 boutons/µm2 x 103, n = 12; p < 0.001; Fig. 6E), and the
number of branch points was also increased ~84% (prec105 = 106.8 ± 7.87, n = 18 versus
B86.1F/Df = 196.2 ± 15.26 branches/µm2 x 106, n = 12; p < 0.001; Fig. 6F). Typically,
branches arise from a single bouton, giving rise to two new synaptic boutons (Zito et al.,
41
1999). In contrast, Synd mutants have an average of ~2.0 hyperbranched boutons per
NMJ or 25.76 ± 2.72 hyperbranched boutons/per muscle area µm2 x 106 (Fig. 6G).
Postsynaptic expression of a Synd transgene rescues the overgrowth phenotype
To prove that the morphologic phenotype was specific to loss of Synd and to
identify the site of action of Synd, we attempted to rescue the NMJ overgrowth
phenotype by tissue-specific expression of Synd using UAS/GAL4 system. Postsynaptic
(muscle) expression of UAS-synd driven by a Gal4 under the control of a myosin heavy
chain promoter (MHC82-Gal4) restored the NMJ architecture to that resembling wild
type larvae as judged by several criteria. NMJ span, bouton number, axonal branching
and hyperbranching phenotypes were rescued. Complete rescue was observed since
numbers of synaptic boutons (prec105 = 1.08 ± 0.07, n = 18 versus muscle rescue = 1.13
± 0.06 boutons/µm2 x 103, n = 12; p > 0.05; Fig. 6E), branching boutons (prec105 =
106.8 ± 7.87, n = 18 versus muscle rescue = 118.3 ± 8.63 branches/µm2 x 106, n = 12; p >
0.05; Fig. 6F), and hyperbranched boutons (prec105 = 1.38 ± 0.95, n = 18 versus muscle
rescue = 5.02 ± 1.80 boutons/µm2 x 106, n = 12; p > 0.05; Fig. 6G) were statistically
similar. Presence of UAS-synd and the GAL4 alleles alone were not sufficient to rescue
(data not shown). Muscle rescue flies hatch in approximately Mendelian ratio and appear
to regain their ability to fly. The only phenotypic parameter not fully rescued was NMJ length. Pan-neural (thus presynaptic) expression of Synd with both Elav-GAL43A4 and
Elav-GAL43E1 drivers also failed to rescue the NMJ overgrowth phenotype in synd
mutants. The ability of muscle, but not neural, expression of Synd to rescue the
morphological phenotype supports the assertion that Synd acts postsynaptically to
42
regulate synaptic development. The successful rescue shows that the NMJ overgrowth in mutants is specifically due to a loss of Synd postsynaptically.
Drosophila Syndapin binds Wsp and the conserved proline487 residue of Synd is required for this interaction
Overlay analysis (Far Western) was used to demonstrate that a GST-Synd-SH3 domain fusion protein binds to recombinant Drosophila Wsp. Extracts from bacteria expressing recombinant Drosophila Wsp were probed with GST-Synd-SH3 or GST alone. Fusion protein binding was detected with an anti-GST antibody (Fig. 4A). GST-
SyndSH3 binding is evident only in cultures where protein expression was induced, but not observed in un-induced cultures. Also, GST alone fails to bind Wsp, thus indicating that this interaction is specific. Within the SH3 domain, a conserved proline residue has been shown to be critical for SH3 binding to proline rich domains. A proline to leucine mutation at residue 434 abrogated binding of mammalian Synd I with Wsp, and an analogous mutation in sem-5, a C. elegans Grb2 homologue, caused a lethal phenotype
(Clark et al., 1992; Qualmann et al., 1999). To address whether the biochemical interaction of Synd and Wsp are dependent on the classical binding interface of the SH3 domain, we mutated amino acid residue 487 from proline to leucine. The P487L point mutation completely abolished the binding capacity of GST-Synd-SH3 to recombinant
Wsp in overlays (Fig. 4A). This evidence further supports the specificity of binding between Synd and Wsp and that this interaction is dependent on the SH3 domain consensus sequence.
43
Syndapin and Wsp interact in vivo
Having shown that Synd and Wsp interact biochemically and share a common
cellular distribution, we wanted to test whether they interact in vivo. If Synd and Wsp
function in a common biological pathway, we predicted that loss of synd would
phenocopy loss of Wsp. We confirmed the NMJ overgrowth phenotype of Wsp loss-of
function shown previously (Coyle et al., 2004), and we found that Synd mutants display
similar degree of overgrowth when compared to loss of Wsp (Fig. 7). The hemizygote
wsp1 over the Wsp deficiency, Df(3R)3450, was used for our analysis. Furthermore, we
reasoned that, progressive loss wild type Synd and Wsp alleles would lead to
progressively worse phenotypes. Hyperbranched boutons are a distinguishing feature of
wsp and nwk mutants (Coyle et al., 2004). NMJs of B86.1F/Df also contain boutons that
branch 3, 4, and even 5 times, which are rarely seen in wild type controls (Fig. 6). In our
hands, wild type NMJs infrequently contain boutons branching to form three additional
boutons, but never contain hyperbranching to four or five boutons (Fig. 6B). We then
turned our attention to determining if synaptic growth is dependent on gene dosage of
Synd and Wsp by monitoring the degree of hyperbranching in synd and/or wsp mutants.
Our most severe allele combination B86.1F,wsp1/B86.1F,wsp1 was lethal as 1st
instar larvae precluding morphologic analysis. B86.1F homozygotes, wsp1 homozygotes
and hemizygous combinations (B86.1F/Df and wsp1/Df) survive till at least pupal stages,
suggesting that lethality is not likely due to second site mutations on the respective
chromosomes. Decreased viability may be due to a decrease in overall health rather than
specifically implicating a common pathway for Synd and Wsp. However, when we examined NMJ morphology phenotypes, there appears to be a trend of increasing severity
44
with decreasing gene dose of wild type synd and wsp. This enhancement was most
striking for the increase in number of hyperbranching boutons, a phenomenon also
observed for nwk and wsp double mutants. This phenotype is greatly enhanced for the
heteroallelic combinations when compared to loss of synd or wsp alone.
B86.1F,wsp1/Df(3R)BSC43,+ showed an ~89% increase in the number of hyperbranched boutons compared to B86.1F/Df(3R)BSC43 (P < 0.001, Fig. 7). Likewise a ~333%
increase in hyperbranched bouton number was observed in B86.1F,wsp1/+,Df(3R)3450
relative to wsp1/Df(3R)3450 (P < 0.001, Fig. 7). Dosage-sensitive genetic interactions are
rarely observed for genes mechanistically unrelated. Taken together, these data suggest
that Synd and Wsp interact in vivo to regulate synaptic development.
Synd acts upstream of wsp
We performed additional immunohistochemical analysis to further investigate the
function of Wsp and Synd at the NMJ. Synd mutants were co-labeled with anti-Wsp and
anti-HRP antibodies to determine if loss of synd would selectively affect Wsp
distribution. At higher magnification images of wild type NMJs, Wsp is seen both presynaptically (colocalizing with anti-HRP staining) and postsynaptically, appearing as a halo around the synaptic bouton (Fig. 8A). In synd mutants, there is a perceptible
decrease in the extent and relative intensity of postsynaptic Wsp, suggesting that
postsynaptic Wsp expression was decreased and/or being mistargeted (Fig. 8B). The
relative level of Wsp immunofluorescence was quantified as a function of relative
distance across the synaptic bouton. Briefly, Z-stack confocal images were acquired and
the optical section corresponding to the center of a given bouton was analyzed. To
45
analyze Wsp fluorescence along relative distance to the synaptic membrane, orthogonal line profiles were acquired through the center of the bouton so that the bouton borders corresponded to 1/3 and 2/3 the length of the line profile (Fig. 8E). The distance of the line profile (acquired in µm) was normalized from zero to 1.0 by piecewise linear interpolation. Based on this protocol, normalized line profile distance 0.33 corresponds the muscle/motorneuron interface and distance 0.66 corresponds to the neuron/muscle border on the opposite side of the bouton. Quantification reveals that relative to wild type controls (prec105) Wsp is decreased in synd mutant most prominently on the postsynaptic side (Fig. 8E). No obvious change in Synd level was observed at the NMJ
of wsp1/Df compared to prec105 (wild type) (Fig. 8C & D). We concluded that Synd
staining is not affected in a wsp mutant because it functions upstream of wsp. This is
supports our hypothesis that Synd binds and recruits Wsp to the NMJ.
46
Figures
Figure 1 - Amino acid sequence alignment of Drosophila Synd with other known
Synd isoforms
The amino acid sequence of Drosophila Synd (gi24648543) was aligned against mouse
PACSIN II (gi7106381), rat Synd IIa (gi6651167), xenopus Synd (gi27696861) and C.
elegans Synd (gi17567725). This was done using GCG’s (Accelrys, San Diego, CA)
pileup and pretty box. Black boxes represent identical amino acids conserved in at least
three of the proteins; gray shaded boxes represent conserved amino acids. Conserved
protein domains have been denoted by bars above the relevant sequence. FCH
(cdc15/FER/CIP4 homology) domain, red bar (first); coiled coil, blue bar (second); NPF
motifs, magenta bar (third to fifth); SH3 (src homology 3) domain; gray bar (sixth).
47
48
Figure 2 - Schematic representation of the gene structure of Drosophila Synd.
(A) A Drosophila Synd homolog has been predicted by sequence annotation of the
genome. The EST LD46328 is a full length clone of Synd. The intron/exon structure was
determined by a Sim4 alignment (Florea et al., 1998) using the LD46328 sequence and
genomic DNA sequences (AE003732). The Synd gene contains ten exons, of which nine
contain the open reading frame; the predicted intron/exon structure of Synd (labeled
Synd, shown in dark gray) is very similar to that of the LD46328 (in light gray) with the
exception of the first exon and 3’ untranslated region not being predicted. The arrow in
Synd represents the start codon of Synd and the pale shading of LD46328 represents the
open reading frame. The open triangles represent the insertion site and orientation of the
transposon (P-element) lines in or around Synd. The imprecise excision line, B86.1F, was generated by mobilization of the EP(3)0877 P-element and is a 536 bp deletion of to the right of the EP(3)0877 insertion site. This deletion removes the entire first exon. (B) This
diagram (to scale) shows the relative positions of the conserved protein domains in
Drosophila Synd. There is an N-terminal FCH domain, a central coiled coil domain, a single NPF motif and a SH3 domain at the very C-terminus of the protein. (C) A western blot of a control line showing that the antibody recognizes a single protein band at 68 kDa, the predicted weight for Synd. Protein was isolated from 3 fly heads. (D) Western blot of protein from a precise Synd excision line, Precise 105, and a Synd mutant,
B86.1F/Df. Protein was extracted from whole adult flies, pooled, and then loaded
according to relative protein amounts. Anti-SNF staining was used to ensure equal sample loading. Synd level in B86.1F/Df is equivalent to ~3% of wild-type, indicating
that B86.1F is a hypomorphic allele.
49
A
B86.1F ( )
B
C D
50
Figure 3 - In wild type embryos and larvae, Synd is distributed in actin-rich, highly
curved structures
(A) Synd is localized in the neuropil of the embryonic nervous system where it
colocalizes with actin, visualized by phalloidin staining. TOTO3, a DNA stain, has been
used to show the position of soma in the CNS. Panel B shows a confocal image of Synd
distribution in the eye 48 hours after pupal formation (APF). A characteristic horseshoe
pattern of staining can be seen in the center of the ommatidia (arrows), the positions of
individual photoreceptors can been as the dark shapes surrounding the Synd staining
(arrowheads). This pattern of staining resembles that of actin, amphiphysin and bifocal.
(C) Synd localization 72 hrs APF. (D) Phalloidin staining 72 hrs APF. Rhabdomeres are clearly visible as puncta in the center of each ommatidium (arrowheads, C and D). (E)
Synd staining (green) in the embryonic hindgut closely colocalizes with phalloidin (red).
Soma in the CNS are labeled with TOTO3 (blue). (F) Synd is prominent in the lumen of the salivary gland and is present at cleavage furrow during early embryogenesis (G-H).
Synd (green) strongly colocalizes with phalloidin (red) during early (I), mid (J), late (K),
and at the end of cellularization.
51
E
Syndapin phalloidin TOTO3 merge F G H
I JF
K L
phalloidin Syndapin merge phalloidin Syndapin merge
52
Figure 4 - Far western blots using GST-Synd-SH3 domains to bind recombinant
Wsp.
(A) A GST-fusion of Synd SH3 binds to bacterially expressed recombinant fly Wsp in overlay assays. Protein extracts from uninduced (u) or induced (i) bacteria were allowed to bind 0.2 µM GST-Synd-SH3 or GST control and then probed with anti-GST antibody.
Fusion protein binding is evident to lysate from induced cultures. Little or no binding is seen with GST or in uninduced bacteria (u) showing that interaction is specific. GST- fusion of a mutant form of Synd SH3 domain fails to bind Wsp suggesting that the proline487 is critical for Synd-Wsp binding. Panel (B) shows coomassie staining of
recombinant Wsp input protein.
53
A B
GST GST-Synd-SH3 GST-Synd-SH3 Coomassie P487L mutant
U I U I U I U I
54
Figure 5 - Localization of Synd at the neuromuscular junction.
(A-F) Synd is found at the neuromuscular junction. There is strong colocalization of
Synd (red) with the post-synaptic marker, discs large (dlg, green) in (F), but a lesser degree of colocalization with the synaptic vesicle marker, cysteine string protein (csp, green) (C). Confocal images were obtained from NMJ 6/7 in third instar larvae. (G)
Wild type NMJ 6/7 were colabeled with Synd (red) and dlg (green). Synd staining is evident in both type Ib and type Is boutons, similar to the distribution of dlg. Synd levels at the NMJ are greatly reduced in the hypomorphic line B86.1F/Df when compared to wild type (wt) controls (H).
55
G H
Syndapin wt
dlg B86.1F/Df
56
Figure 6 - Loss of postsynaptic Synd leads to altered neuromuscular junction
morphology
Confocal images of NMJ 6/7 (A-C) co-stained with anti-HRP (green) and anti-Syt (red).
Top panel (A) shows the typical appearance of wild-type NMJs. The Synd hypomorph
B86.1F/Df displays a pronounced NMJ overgrowth phenotype. Expression of Synd cDNA in the muscle (C), but not neurons, of mutant larvae rescues the morphologic phenotype. Scale bar equals 20 µm. (D-G) Quantification of synaptic morphology parameters. NMJ length (D), number of synaptic boutons (E), number of branching boutons (F), and number of hyperbranching boutons (G) for NMJ 6/7 were normalized to muscle 6/7 area. Neuronal expression of a Synd cDNA fails to rescues the Synd mutant phenotype indicating that postsynaptic Synd is critical for regulating synaptic morphology. Genotypes are as follows: UAS-Synd/+ ;; MHC82-Gal4, B86.1F/Df for muscle rescue, UAS-Synd/+ ;; elav3A4-Gal4, B86.1F/Df for neuronal rescue 1 and UAS-
Synd/+ ;; elav3E1-Gal4, B86.1F/Df for neuronal rescue 2. (mean ± S.E.M; * indicates
significant difference from precise 105, p < 0.05; ** indicates p < 0.01; and ***
indicates p < 0.001)
57
58
Figure 7 - Gene dosage of Synd and Wsp affects hyperbranching phenotype
Quantification of mean number of branching synaptic boutons giving rise to three or more boutons. The number of hyperbranching boutons of NMJ 6/7 was normalized to corresponding muscle 6/7 area. Partial loss of wild type Wsp
(B86.1F,wsp1/Df(3R)BSC43) enhances the hyperbranching phenotype of a Synd mutant
(B86.1F/Df(3R)BSC43). Likewise, the loss of Synd (B86.1F,wsp1/Df(3R)3450) exacerbates the phenotype of a Wsp mutant (wsp1/Df(3R)3450). The number of hyperbranches for loss of one copy Synd (B86.1F/+), Wsp (wsp1/+), or both
(B86.1F,wsp1/+) was not significantly different from precise excision wild type (prec
105, +). (mean ± S.E.M; * indicates significance, p < 0.001, ANOVA).
59
60
Figure 8 - Post-synaptic Wsp is decreased in a Synd mutant
Confocal images of synaptic boutons in wild type (wt) (A) and Synd mutant (B86.1F/Df)
(B) colabeled with anti-Wsp (green) and anti-HRP (red). Images were generated from
single optical slices (.22µm). Postsynaptic Wsp is affected in Synd mutants evident by
the change in extent of Wsp staining relative to the neuronal membrane (delineated by
HRP). Scale bar is 2 µm. Loss of wsp does not affect Synd expression. Panels (C) and
(D) show projected confocal images of NMJ 6/7 from wild type (wt) and Wsp mutant
(wsp1/Df) larvae, respectively. There is no significant change in Synd levels when Wsp is
lost. (E) Line profile analyses of confocal images from NMJ of wild type (prec105) and
Synd mutant (B86.1F/Df) larvae colabeled with anti-Wsp and anti-HRP. Line profiles
were generated by marking synaptic bouton borders at 1/3 and 2/3 the length of the line
profile. Dashed vertical lines signify the neuronal membrane, relative distances 0.33 -
0.66 represents presynaptic compartment and distances 0 - 0.33 and 0.66 - 1.0 represent postsynaptic compartments. Borders of the synaptic bouton are revealed by peaks of
HRP staining (dark blue, n = 20) and Wsp staining across synaptic boutons from wild type (prec 105, red, n = 10) and Synd mutant (B86.1F/Df, green, N = 10) are shown.
Post-synaptic Wsp is lost in a Synd mutant (green) relative to wild type (red). Intensity was normalized to maximum intensity of given line profile. Distance was normalized to the total distance across the midsection of a single synaptic bouton. Black arrows indicate edges of the synaptic bouton and area to the left and right of the arrows signifies post- synaptic distribution. Taken together, this data suggests that Synd acts upstream of Wsp.
61
wt
62
CHAPTER 3
DISCUSSION
63
Research Conclusions
In this study, we describe the cloning and characterization of Drosophila Synd
(Synd). We have shown that Synd is required at the NMJ to regulate proper synaptic growth. We believe that Synd is necessary for the efficient recruitment of Wsp to the
postsynaptic plasma membrane, thereby regulating synapse formation and remodeling.
The following lines of evidence have led us to this conclusion. First, Synd and Wsp
interact biochemically and this interaction is dependent on the canonical SH3 binding
motif. Second, Synd and Wsp are both found in the muscles (postsynaptic compartment) along the NMJ. Third, synd mutants show a neuromuscular junction overgrowth phenotype that phenocopies the loss of Wsp. Fourth, synd and wsp genetically interact in a dosage-sensitive manner, where loss of one copy of wild-type wsp enhances the phenotype observed in a synd mutant. Finally, postsynaptic Wsp is lost in a synd mutant.
Taken together, our data indicate that synd is a postsynaptic Wsp adaptor that governs
Drosophila NMJ architecture.
Postsynaptic actin cytoskeleton and the control of synaptic morphology
Regulators of synaptic growth have been identified in mutants with aberrant NMJ
morphology. It has been suggested that synaptic activity, protein turnover, cell adhesion
molecules, and modulators of microtubules are involved in this process (Budnik et al.,
1990; Ruiz-Canada and Budnik, 2006; Schuster et al., 1996; van Roessel et al., 2004).
Although some of the signaling pathways that control synapse growth [e.g. JNK, AP-1,
cAMP and retrograde bone morphogenetic protein (BMP) signaling pathways] have been
well described, it is only recently that effectors of cell shape have been clearly implicated
64
(Etter et al., 2005; Marques and Zhang, 2006; Sanyal et al., 2002). Recent work suggests
that a protein called nervous wreck (Nwk) interacts with Wsp in a signaling complex that
regulates synaptic growth, thus also implicating the actin cytoskeleton into the control of
synaptic growth (Coyle et al., 2004). However, Nwk is only found presynaptically, while
Wsp is found both pre- and postsynaptically, and presynaptic expression of Wsp only
leads to partial rescue of the overgrowth phenotype. This implies that either nwk acts
presynaptically independent of wsp, that presynaptic wsp expression was insufficient for
full rescue, or that Wsp expressed in other cell types is necessary for correct synaptic
development. Thus, the question becomes: what acts as to localize Wsp on the
postsynaptic side of the synapse? We believe Synd fulfills this role. Our data shows that
Synd is enriched postsynaptically at the neuromuscular junction and is a key molecule for
the recruitment and/or stabilization of Wsp to the post-synaptic membrane of the fly
NMJ. The ability of a synd transgene to rescue the NMJ overgrowth phenotype by
expressing postsynaptically in the muscle, but not presynaptically by a pan neural driver,
strongly argues that postsynaptic Synd is required for modulating synaptic growth.
One distinguishing feature of the nwk and wsp overgrowth phenotypes is the
presence of hyperbranched boutons. Our data show that Synd mutants also contain
hyperbranched boutons. Taken together these data imply that regulators of the Wsp/actin
on both sides of the synapse are most important for inhibiting synaptic branching. This
further implicates the functional role of the actin cytoskeleton in the postsynaptic target
cell in determining synaptic morphology.
65
Syndapin is an adaptor protein that recruits Wsp to the postsynaptic membrane
Syndapins contain an N-terminal FCH (FER/CIP4-homology) domain and coiled- coil domain that are collectively referred to as the BAR or F-BAR
(Bin/Amphiphysin/Rvs-homology) domains (Itoh et al., 2005). Structural and functional analysis of BAR domains have shown that they assemble as crescent-shaped dimers that can bind cell membrane by virtue of positively charged residues (Dawson et al., 2006;
Peter et al., 2004). Purified BAR domains are able to bind and tubulate liposomes in vitro and plasma membrane in cell culture (McMahon and Gallop, 2005; Peter et al.,
2004; Razzaq et al., 2001). The biological functions of BAR domain containing proteins are believed to involve membrane association or development of stable membrane curvature. Like other BAR domain family members, mammalian Syndapin I is able to form homodimers and tetramers (Halbach et al., 2007; Kessels and Qualmann, 2006). In addition, recombinant Syndapin I is able to tubulate liposome in vitro (Itoh et al., 2005).
We hypothesize that direct plasma membrane binding may enable Synd to recruit Wsp to the NMJ.
Having both Synd and wsp mutant alleles and the respective antibodies, we are able to see which protein is lost in which mutant background. Furthermore, our data indicate a dose-dependent genetic interaction, where loss of both genes are more severely affected than either single mutant. This highlights the mechanistic connection between synd and wsp, but suggests that Synd has important functions that effect NMJ development beyond Wsp recruitment. It would not be surprising if Synd recruits other molecules to the NMJ. Indeed, the regulation of synaptic growth is likely to involve
66
many signaling molecules and effectors, and the importance of postsynaptic association
of Synd and Wsp may become apparent for other synaptic functions.
A putative role for Drosophila Syndapin in synaptic vesicle endocytosis
Studies performed by the Ramaswami group suggest that "syndapin is dispensable
for synaptic vesicle endocytosis" by reporting that a line with a transposon insertion in the synd 3' UTR has decreased Synd, but normal synaptic transmission (Kumar et al.,
2009a). However, using more severe hypomorphs than those previously used, our lab has
generated preliminary data suggesting that synd may in fact regulate SVE. This suggests
a presynaptic role for synd. Under high frequency stimulation, there is a progressive
decrement in evoked potential over time (data not shown). This resembles the phenotype
of other endocytosis mutants like endophilin and is indicative of defective SVE and
failure to properly replenish (Verstreken et al., 2003). This rundown in evoked release
for both synd and endophilin reach a plateau phase, unlike that seen in dynamin (shibire)
mutants (Verstreken et al., 2003). Driving transgenic expression of synd in the neuron is
able to rescue the "run-down" phenotype, and interestingly, there is partial rescue when
the transgene is expressed by a muscle driver. The significance of postsynaptic effects on
SVE and synaptic transmission is unclear, but may involve a compensatory structural
change induced by muscle Synd, modification to the postsynaptic density and then
perhaps glutamate receptor density. Further study is needed to clarify the roles of pre
and/or postsynaptic Synd in the regulation of transmission. On the surface, it is
enigmatic how muscle expression of a protein could rescue a presynaptic defect. To
further investigate this, immunofluorescent staining of Synd, Wsp, and endocytosis
67
proteins such as endophilin, amphiphysin, and dynamin in mutant and rescue flies would
be illustrative in describing how postsynaptic expression of Synd may have direct or
indirect influences on SVE machinery. Also, examination of the postsynaptic density
(discs large/PSD-95) and glutamate receptor density and distribution in mutants and
rescue flies may explain this phenomenon.
Remaining questions and future directions
The supernumerary bouton phenotype observed in synd, nwk and wsp mutants has also been observed at the NMJ of endocytosis mutants (Dickman et al., 2006; Koh et al.,
2007; Marie et al., 2004; Sweeney and Davis, 2002). This common morphological phenotype suggests that endocytic mechanisms and retrograde BMP signaling are linked in a common regulatory pathway to influence structural development. These observations indicate a clear presynaptic (neuronal) function for the generation of hyperbranched boutons at the NMJ. For Nwk, this can now be explained by its modulation of BMP signaling by binding to the type I receptor, Tkv (O'Connor-Giles et al., 2008). Nwk suppresses BMP signaling by presumably regulating the endocytosis of ligand-activated BMP receptors.
Since synd mutation also produces a hyperbranched bouton phenotype. Synd could be acting by an novel, independent mechanism, or perhaps like nwk, synd may regulate endocytosis and/or BMP receptor trafficking to affect morphogenesis. However, because synd acts postsynaptically to regulate synaptic growth, hypothetical integration of synd into the framework of other known presynaptic modulators of synaptic morphology is more difficult. One possibility would be the regulation of Gbb (BMP
68
ligand) release from the muscle, thus regulating BMP signaling. Very recently, a
postsynaptic regulator for synaptic growth has been identified. Drosophila Cdc42-
interacting protein 4 (dCIP4) has been found to act downstream of Cdc42 to activate Wsp
in the muscle. Wsp-Arp2/3 activation then inhibits Gbb secretion, therefore dCIP4 acts to suppress Gbb release (Nahm et al., 2010). It is very plausible that Synd recruits Wsp
postsynaptically, which would facilitate inhibition of Gbb secretion. In synd mutants, the
failure of Wsp localization could then result in excess Gbb release and excess downstream BMP signaling. Postsynaptic synd regulation of BMP signaling could be examined in both synd mutant and muscle rescue lines by looking at levels of phosphorylated Mothers against Dpp (MAD). To reiterate, MAD is an R-Smad that is
phosphorylated in response to of type I and II BMP receptor dimerization (Nahm et al.,
2010; Rawson et al., 2003). If Synd regulates Gbb release, then phospho-MAD staining
should be elevated. Another approach would be the use of a fluorescently labeled Gbb
ligand to track efficiency of secretion. Transgenic expression of Gbb-GFP both in wild
type and mutants would serve as a visual marker for Gbb release. Finally, if transgenic
expression of BMP suppressor genes acting downstream of synd (wsp, nwk and hiw) rescue the loss of synd phenotype, then these data would strongly validate this model and implicate postsynaptic synd as a regulator of Gbb release.
Synd mutants and rescue lines could be further examined for staining of other endocytosis and BMP signaling proteins. We have already shown that loss of synd impairs Wsp accumulation at the synapse. However, if and how synd affects the endocytic machinery or proteins of the BMP pathway has not been established. Going further, it would be of great interest to uncover further genetic and molecular interactions
69
between synd and other morphogenic pathways such as Wg/Wnt signaling. These studies
would not only clarify the molecular mechanisms underlying Synd mediated plasticity, but may uncover additional relationships between the known modulators of synaptic morphology.
More generally, it would be interesting to determine the mechanistic relationship
between SVE and regulation of synaptic growth by "endocytosis genes." Is it possible for these dynamin interacting, F-BAR and SH3 containing proteins to have an distinct presynaptic (SVE) and postsynaptic (synapse growth) roles? Or does changes in endocytosis dynamics modulate the localization of BMP receptors thereby affecting signaling and morphology? Cell type specific (pre- versus postsynaptic) rescue
experiments for most of the endocytosis mutants have not been performed and would be
extremely insightful. Furthermore, immuno-gold labeling of Gbb and BMP receptors in
ultrastructural studies of endocytosis mutants would highlight receptor distribution and
Gbb secretion/accumulation. These future studies would start build a predictive framework about the principles and mechanisms guiding morphologic plasticity at the synapse.
70
CHAPTER 4
INTRODUCTION TO PART II
71
The function and importance of the serotonergic system
Serotonin (5-hydroxytryptamine; 5-HT) is a monoamine extracellular signaling molecule that is biochemically synthesized from tryptophan. In the central nervous system, it is released by a small number of neurons mainly restricted to the ventral regions of the hindbrain. The hindbrain develops into the cerebellum, pons, and medulla surrounding the fourth ventricle, and in the adult nervous system serotonergic neurons are located in the raphe nuclei that are restricted to the basal plate of the pons and medulla
(Kandel et al., 2000). Additionally, neurons with serotonergic phenotype are found in the fetal hypothalamus, adult dorsomedial hypothalamic nucleus and the in the postnatal spinal cord. From the rostral raphe nuclei, serotonergic neurons project their axons into the midbrain and forebrain. The majority of these fibers ascend into the cerebral cortex.
The caudal raphe nuclei projects descending fibers into the spinal cord, where they innervate the intermediolateral column (preganglionic sympathetic neurons), somatic motor neurons and dorsal horn neurons (Adell et al., 2002; Rubenstein, 1998). Thus, serotonergic neurons project into higher order brain regions, and influence the autonomic and peripheral nervous system.
Serotonin regulates numerous physiological functions and behaviors and is involved in the pathogenesis of several disease processes. Serotonin has been demonstrated to play important roles for controlling among others: mood, sexual behavior, feeding, sleep/wake cycle, memory, cognition blood pressure regulation and breathing (Mooney et al., 1998). In a clinical context, serotonin dysfunction has been related to neuropsychiatric disease process such as depression, anxiety, schizophrenia, obsessive compulsive and autism spectrum disorders. (Davidson et al., 2000; Lucki,
72
1998; Mann et al., 2001; Nelson and Chiavegatto, 2001; Pardo and Eberhart, 2007).
Serotonergic system defects have even been linked to cardiovascular disease and respiratory disorders such as SIDS susceptibility and asthma and (Barnes et al., 1998;
Nebigil et al., 2001; Richerson, 2004). Therefore, understanding the role of serotonergic function has been the focus of intense research and is important for understanding human physiology, behavior and disease.
Genetic studies in have been informative as to the describing the necessity of functional serotonin circuits in mammals. Although genetic ablation of serotonergic neurons in animal models does not cause primary lethality, mutations affecting the 5-HT system cause behavioral and physiological abnormalities which ultimately impair survival (Ding et al., 2003; Hendricks et al., 2003; Zhao et al., 2006). For example, mice lacking the pheochromocytoma 12 ETS factor-1 (Pet-1) promoter that is required for 5-
HT neuron specification and development, lack the majority of their serotonergic neurons. These mice display increased anxiety behaviors and are more aggressive
(Hendricks et al., 2003). Furthermore, females fail to display instinctive nurturing behaviors such as huddling pups and nest building, impacting the survival of offspring
(Lerch-Haner et al., 2008). In addition, newborn Pet-1 (-/-) mice have depressed breathing frequency and a higher incidence of spontaneous and prolonged respiratory pauses (Erickson et al., 2007). Breathing defects are also observed in mice that lack almost all serotonergic neurons caused by a knocking out Lmx1b (LIM homeobox transcription factor 1 beta) in Pet-1 expressing neurons (Hodges et al., 2008; Zhao et al.,
2006). These mice have delayed respiratory maturation and aberrant response to hypercapnia. The Lmx1bf/f/p mice exhibit impaired thermoregulatory responses to cold
73
ambient temperatures without affecting thermosensory perception or heat conservation by
peripheral vasoconstriction (Hodges et al., 2008). The synthesis of serotonin from
tryptophan in mammals depends on the activity of tryptophan hydroxylase (TPH)
isoforms. Genetic ablation of the CNS isoform, Tph2, causes marked growth retardation,
decreases postnatal survival and impairs respiration and thermoregulation (Alenina et al.,
2009). Behaviorally, Tph2 (-/-) mice are more aggressive and females display maternal neglect like the Pet-1 (-/-) mice (Alenina et al., 2009). Others also observe depression- like behaviors in Tph2 null mice, which are enhanced when Tph1 is also deleted
(Savelieva et al., 2008). Taken together, the evidence from genetic ablation studies illustrate the importance of serotonin neuron activity for normal behavior, reproductive success and survival.
Serotonin exerts it action by binding to and activating metabotropic or ionotopic
5-HT receptors. 5-HT receptors are expressed in neurons of the central, peripheral and enteric nervous systems and are expressed in several other tissue types including the gut, cardiovascular, pulmonary systems and blood (Cazzola and Matera, 2000; Hoyer et al.,
2002). In humans, 5-HT receptors are organized into seven subfamilies (i.e. 5-HT1-7) and all receptors with the exception of 5-HT3 are G protein-coupled receptors (GPCRs).
Within these seven subgroups, several variants and isoforms can be found (Barnes and
Sharp, 1999; Hoyer et al., 2002; Kroeze et al., 2002). The 5-HT3 receptor is a ligand-
gated ion channels (ionotropic), which triggers rapid depolarization mediated by high Na+
and Ca2+ permeability. The 5-HT GPCRs can be divided into three major subgroups depending on which pathway they activate. 5-HT1 receptors couple mainly to the Gi/o
74
pathway, 5-HT4, 5, 6, 7 couple to the Gs pathway, while 5-HT2 receptors activate the Gq
pathway.
The 5-HT1A receptor and its role in disease pathogenesis and treatment
Five 5-HT1 receptor subtypes have been identified to date (5-HT1A, 1B, 1D, 1E, 1F).
Of these, the 5-HT1A receptor, has been identified to be an important regulator of 5-HT signaling. 5-HT1A receptors are distributed widely throughout the brain, but are particularly enriched in the limbic system structures (Lanfumey and Hamon, 2004). 5-
HT1A activation induces hypothermic responses, decreases blood pressure and heart rate
and increases locomotor responses (Dreteler et al., 1991; Kalkman, 1995). In the dorsal
raphe, the largest serotonergic nucleus, 5-HT1A is found in cell bodies and dendrites of
neurons (Riad et al., 2000; Sotelo et al., 1990). These receptors are involved in the auto-
regulation of 5-HT release, where their activation leads to a decrease in firing via
hyperpolarization of the cell membrane (Stamford et al., 2000). The major effects of 5-
+ HT1A receptor activation are inhibition of adenylate cyclase and induction of K currents
(Colino and Halliwell, 1987; Weiss et al., 1986). After ligand binding and subsequent
Gi/o coupling, dissociated βγ subunits bind to and increase the permeability of G protein-
regulated inward rectifying potassium channels (GIRK) (Kofuji et al., 1995). This results
in increased K+ efflux and membrane hyperpolarization. Thus, 5-HT input onto
serotonergic neurons causes a decrease in 5-HT neuron firing via 5-HT1A activation. 5-
HT1A function has been implicated in a wide variety of physiological and behavioral
functions including learning and memory, thermoregulation, neuro-endocrine regulation,
sexual behavior, food intake, immune function, and aggression (Leone et al., 1998;
75
Raymond et al., 2001; Seletti et al., 1995). Furthermore, 5-HT1A has been implicated in
the pathogenesis of anxiety and depression and is thought to be a physiologic target for
selective serotonin reuptake inhibitor (SSRI) treatment (Lanfumey and Hamon, 2004).
Although to date it has been one of the most extensively studied of the serotonin
receptors, there is still much ambiguity regarding the normal in vivo function of 5-HT1A
and dysfunction in the context of disease. Several independently generated 5-HT1A
knockout mice reveal increased anxiety as demonstrated by elevated maze, open field,
and novelty-suppressed feeding tests, as well as insensitivity to 8-OH-DPAT (agonist)
and WAY100365 (antagonist) administration (Ase et al., 2001; Heisler et al., 1998; Parks
et al., 1998; Ramboz et al., 1998). Furthermore, these mice exhibit a marked
antidepressant-like response in tail-suspension assays meaning that the effect of gene
deletion mimics the effects of antidepressant drug administration in behavioral assays
(Heisler et al., 1998; Steru et al., 1985). Serotonin neurons in 5-HT1A null mice also fire
at an increased rate compared to wild type controls (Richer et al., 2002). This may explain the observation that serotonin concentrations in the frontal cortex and hippocampus are increased in knockout mice (Parsons et al., 2001). On the other hand, transgenic expression of 5-HT1A receptors decrease anxiety-like behaviors and males have a lower body temperature, consistent the responses seen with 5-HT1A agonist
administration into wild type mice (Kusserow et al., 2004). Additionally, 5-HT1A
overexpression in serotonin neurons results in increased postnatal lethality and autonomic
dysfunction marked by hypothermia and bradycardia (Audero et al., 2008). Therefore, it
is clear that normal levels of 5-HT1A activity are necessary to produce normal behavior
and physiological control.
76
Identification of receptor-selective agonists, such as 8-hydroxy-2-(di-N- propylamino)tetralin (8OH-DPAT), have been valuable for defining 5-HT1A receptor
function. Systemic application of the 8OH-DPAT induces hypothermia, decreases blood
pressure and heart rate, and increases locomotor response (Hoyer et al., 2002). These
responses are consistent with the transgenic mouse data (Audero et al., 2008; Kusserow
et al., 2004). Generally speaking, 5-HT1A agonists are considered to be anxiolytic
(Gordon and Hen, 2004). Antagonists such as WAY100635 increase anxiety-like
behaviors, although some responses are confounded by interaction with multiple
receptors by a single compound due to limited pharmacologic selectivity (Ramboz et al.,
1998). From a simplistic viewpoint, it would appear paradoxical that 5-HT1A mediated auto-regulation may underlie the 2-4 week latency of selective serotonin reuptake inhibitors (SSRI). That is, drug treatment efficacy marked by increases in 5-HT release in would occur only after 5-HT1A auto-receptor desensitization (Blier et al., 1987). This
would presumably mean that SSRI treatment causes a decrease in 5-HT1A signaling. This
notion is supported by some reports that show that co-administration of 5-HT1A
antagonists accelerates the onset of SSRI action (Artigas et al., 1994; Hjorth, 1993;
Hjorth et al., 2000). Constitutional 5-HT1A KO animals do not exhibit SSRI-induced
changes in behavior (Mayorga et al., 2001; Santarelli et al., 2003). However, SSRI
treatment induces greater increases of serotonin release in the frontal cortices and
hippocampi of KO mice when compared to wild type (Parsons et al., 2001). These sets of data seem at odds with each other since behaviorally, KO mice are resistant to SSRI treatment, but respond with greater increases in serotonin tone. This discrepancy may be
due to the fact that mutant mice are in an anti-depressed state prior to SSRI
77
administration and the increases in serotonin do not manifest a statistically significant
change in behavior. To delineate the specific contribution 5-HT1A function in
serotonergic (autoreceptor), and not peripheral nor non-5HT neurons (heteroreceptor),
Richardson-Jones and colleagues have recently used a novel tetO-based strategy to
decrease 5-HT1A receptor levels only in serotonergic neurons, postnatally. This
conditional knockdown of 5-HT1A autoreceptor expression in 5-HT neurons imparts SSRI
sensitivity to a genetic strain that is normally resistant to drug treatment (Richardson-
Jones et al., 2010). These studies suggest that more subtle variations in serotonin tone
rather than complete absence or presence sets the stage for treatment response. However,
what obscures the interpretation of the interplay between 5-HT1A receptor and SSRI
treatment is the apparent experimental contradictions of wild type mice responses to
drugs. The wild type mice used by Richardson-Jones et al. are refractory to SSRI treatment, but using the same behavioral test in 2003 (novelty suppressed feeding), the
same group found that wild type mice respond to long-term administration of
antidepressant, including fluoxetine (Santarelli et al., 2003). Albeit, experimental
variations with respect to sensitivity to SSRI and basal antidepressed states can occur and
can be explained by individual animal variation and nuances in the execution of
behavioral assays. However, these discrepancies illustrate the complex nature of
delineating the role of 5-HT1A function in the pathogenesis and treatment of disease.
Aside from anxiety and depression, 5-HT1A receptors serve as functional drug
targets for treating schizophrenia, sleep disorders, obesity, and neurodegenerative
diseases among others (Bantick et al., 2001; Francis, 1996; Muraki et al., 2004; Schechter
78
et al., 2005; Vickers and Dourish, 2004). It is therefore of particular interest how 5-HT1A
receptor activation regulates the serotonergic system.
Structural determinants of 5-HT receptor function and targeting
GPCRs consist of 7 transmembrane domains (TMs), which are connected via
intra- and extracellular protein domains. The TMs and the extracellular protein domains
including the N-terminus are needed for ligand binding and conformational shifts after receptor activation. The intracellular protein domains of GPCRs--three intracellular
protein loops [connecting TM1-2 (i1 loop), TM3-4 (i2 loop), TM5-6 (i3 loop)] and the C-
terminal domain (CT)--contain the binding and interaction sites necessary for G protein pathway selectivity and pathway localization. G protein selectivity in 5-HT receptors
involves intracellular loops i2, i3 and the CT (Wess, 1997). Chimeric and mutagenesis
studies on 5-HT receptors and others point in particular to the i3 and the CT for
determining G protein coupling/selectivity, receptor trafficking and interaction with other
intracellular protein of the signaling cascade. In contrast, the very short i1 loop and also
the i2 loop do not seem to play major roles for the selectivity and activation of the 5-HT receptors (Kroeze et al., 2002). The involvement of i3 loop for G protein coupling and selectivity has been shown for 5-HT1A, 1B, 2A, 7 receptors (Egan et al., 1998; Malmberg and Strange, 2000; Obosi et al., 1997; Oksenberg et al., 1995; Pauwels et al., 1999;
Shapiro et al., 2002). The i3 of 5-HT1A also contains PKC phosphorylation sites (Lembo
and Albert, 1995). For 5-HT receptors, the CT is prominently involved in receptor
trafficking by determining the axonal and somatodendritic localization of the receptor
(Darmon et al., 1998; Jolimay et al., 2000). For example, in 5-HT2A receptors the CT
79
contains a PDZ binding domain important for receptor clustering in dendrites (Xia et al.,
2003a; Xia et al., 2003b). For 5-HT1A and 5-HT1B, the CT contains targeting motifs that
are involved in determining somatodendritic versus axonal trafficking, respectively
(Darmon et al., 1998). Interestingly, the divergent targeting pattern for 5-HT1A and 5-
HT1B in neurons is also observed when these receptors are ectopically expressed in an
epithelial cell line, LLC-PK1. 5-HT1A traffics basolaterally in these cells, whereas 5-
HT1B is found apically (Darmon et al., 1998). The CT of 5-HT1A contains 17 amino acid
and contains a di-leucine motif (I414 and I415) and two cysteine palmitoylation sites
(C417 and C420) (Carrel et al., 2006). Di-leucine motifs have been implicated in
transport to the cell membrane (Schulein et al., 1998), endocytosis and exocytosis (Kasai
et al., 2008; Mason et al., 2008), lysosomal targeting (Letourneur and Klausner, 1992)
and sorting to the basolateral membrane (Doumanov et al., 2006; Li et al., 2007).
Palmitoylated cysteine residues have also been suggested to regulate membrane targeting,
but also may be involved in modulating GPCR signal (Qanbar and Bouvier, 2003). CT truncation mutations of 5-HT1A causes the protein to be sequestered in the endoplasmic reticulum (ER) (Jolimay et al., 2000). In fact, mutation of the di-leucine motif alone is sufficient to abrogate membrane trafficking in both cultured cells (LLC-PK1 and COS-7)
and hippocampal neurons (Carrel et al., 2006). Furthermore, exchanging the CT of 5-
HT1A receptor with that of the closely related 5-HT1B receptor results in a chimeric
receptor that is not transported out of the ER (Jolimay et al., 2000). Thus, there are
sequences specific to the CT of 5-HT1A that allow proper targeting to the basolateral
aspect in LLC-PK1 cells and the soma and dendrites of neurons. More recently, Yif1B, a
mammalian ortholog to Yif1p, was identified to be a binding partner to the CT of 5-HT1A
80
using a yeast two-hybrid screen (Carrel et al., 2008). Yif1p is a yeast protein previously
implicated in vesicular trafficking between the ER and Golgi apparatus (Spang, 2004).
The specific interaction between the 5-HT1A CT and Yif1B is necessary for specific targeting to neuronal dendrites. Inhibition of Yif1B by siRNA impairs dendritic trafficking of 5-HT1A. Furthermore, co-expression of the CT of 5-HT1A alone inhibits the
ability of full length 5-HT1A to target to distal segments of dendrites, presumably by
competitively inhibiting the interaction between endogenous Yif1B and full length 5-
HT1A (Carrel et al., 2008).
G protein-coupled receptors of visual systems as "optogenetic" probes
Light sensors are critical for survival as they control behavioral responses such as
light tracking for single-celled organisms and allow for function of visual systems in
higher order animals. These light-sensitive receptors consist of metabotropic G protein-
coupled receptors (GPCRs) that activate downstream signaling pathways and also ion
channels that are gated by light stimulation. Recently, several groups have successfully expressed these light-sensitive molecules in heterologous cell types and have shown that these non-visual cell types can be rendered sensitive to light. This method has been dubbed "optogenetics," and it has been applied with great success to the manipulation of non-photoreceptor neurons. Indeed, a major challenge in neuroscience has been the accurate correlation of activity of specific neurons and circuits to the complex functional outputs of the brain, namely behavior and regulation of physiological processes. For this to occur, not only do neuronal circuits have to be anatomically defined, but specific neurons within intact circuits must be made accessible and then precisely manipulated
81
while leaving the relevant components of circuits intact. Optogenetic approaches have
the potential to accomplish these goals since light-activated proteins can be expressed in a cell specific manner by genetic manipulations, the stimulus (light) is completely absent inside an intact brain, and the stimulus can be controlled on a millisecond time scale.
Thus, the emerging field of optogenetics provides exciting new tools to precisely study the connections between neural function and behavior in intact, behaving animals. In this section of the dissertation, I will review the light-sensitive GPCR candidates for use as optogenetic probes and discuss the feasibility of their use.
The visual systems of invertebrates and vertebrates depend on light-sensitive
GPCRs for the sensation of light. These GPCRs are tuned to respond maximally to specific wavelengths of light along the visible spectrum and activate molecular cascades leading to changes in photoreceptor firing. Phototransduction is initiated by the absorption of a single photon by a retinal aldehyde of vitamin A that is bound to a critical lysine residue in the rhodopsin apoprotein. Light causes photoisomerization of retinal from 11-cis to the all-trans conformation, which in turn leads to dissociation of the G protein from the GPCR and subsequent activation of downstream signaling pathways.
Phototransduction in vertebrates and invertebrates differ in two fundamental ways. The transduction cascades employ divergent G protein pathways, hence signaling involves different enzymes and different second messengers. The selective activation of G protein subtypes by invertebrate (Gq) and vertebrate (Gi/o) rhodopsin is dictated by binding
specificity to α-subunits and not by relative availability of different G-protein subtypes
(Terakita et al., 1998). Furthermore, photoreceptors respond with opposite electric polarity to stimulation. An additional notable difference between vertebrates and
82
invertebrates is the cellular structure of photoreceptors. Photoreceptors of vertebrates are ciliated whereas photoreceptors of invertebrates are organized in a rhabdomeric fashion.
Invertebrate rhodopsin
For invertebrate rhodopsins, light stimulation induces Gq/11 protein activation, which in turn stimulates phospholipase C activity. This leads to production of inositol
(1,4,5)-trisphosphate (IP3) and diacylglyercol (DAG). These second messengers activate non-specific cation channels, which depolarizes the cell. Arrestin binding terminates the
GPCR signal and initiates the biochemical regeneration of 11-cis retinal, from the spent photoligand, all-trans retinal (Kiselev and Subramaniam, 1994). In 2002, Zemelman and colleagues first acknowledged the potential for using invertebrate light receptors to depolarize other non-visual cell types (Zemelman et al., 2002). First, a functional receptor signaling cascade was reconstructed in Xenopus oocytes by the co-expression of ten different proteins. Of these, three components of the invertebrate GPCR (NinaE) signaling cascade were deemed necessary and sufficient for heterologous reconstitution of light sensitivity. These three proteins were the light-activated GPCR (NinaE), the G protein αq subunit and arrestin-2. Heterologous expression of these three proteins and a membrane bound GFP (collectively named chARGe) in hippocampal neurons was accomplished by co-transfection of two separate plasmids and resulted in the induction of action potential firing (spiking) after light activation (Zemelman et al., 2002).
The use of NinaE for the control of neuron firing was initially an exciting development, however, this approach had some practical limitations. First, induction of neurons firing and deactivation were relatively slow and inconsistent making it difficult
83
to precisely control the firing pattern of the neuron. In addition, the precise cellular mechanism by which Gq signaling induced neuron firing was not defined. Thus, the
application of chARGe could possibly be limited to neurons that are endogenously
regulated by Gq pathways. However, what prohibits the practical application of NinaE is
that several additional signaling components must be co-expressed making it difficult to
express by viral vectors or by transgenes in live animals.
Vertebrate rhodopsins
Light stimulation of vertebrate rhodopsin induces isomerization of a vitamin A
derivative, which then leads to a series of intramolecular changes and the formation of
the active form for rhodopsin (Metarhodopsin II). Conversion to metarhodopsin and
activation of G protein signaling occurs within a few milliseconds of light exposure
(Dickopf et al., 1998). The retinal chromophore is covalently bound to a lysine 296
within the seventh transmembrane of rhodopsin through a protonated Schiff base
(Herlitze and Landmesser, 2007). Photoisomerization of 11-cis retinal causes a series of
conformational changes within the protein and activates the G protein, transducin,
belonging to the Gi/o subfamily. Transducin thus activates phosphodiesterase, which
catalyzes the hydrolysis of cyclic guanosine monophosphate (cGMP) to 5-GMP. The reduction in cGMP causes cGMP-gated cation channels to close, which reduces Na+ and
Ca2+ influx. The resulting cell membrane hyperpolarization decreases action potential
firing (Ebrey and Koutalos, 2001; Fain et al., 1996). Because of intracellular signal
amplification, it has been estimated that a single photoisomerization event leads to a
change in the influx of at least 105 cations illustrating exquisite sensitivity of
84
photoreceptors. In addition to 11-cis retinal, the binding site of rod rhodopsin can also
accommodate other retinal isomers (e.g. 9-cis and 13-cis), though the activation characteristics of these chromophores differ (Han and Sakmar, 2000; Kefalov et al.,
1999).
In neurons, activation of Gi/o signaling leads to a reduction in action potential
firing rate by activating G protein-coupled inward rectifier K+ channel (GIRK), which
+ causes K efflux and hyperpolarization. Additionally, Gi/o activation presynaptically can
inhibit Ca2+ channels and subsequent synaptic vesicle release. Because the α subunit of
transducin belongs to the Gi/o subfamily, our lab hypothesized that vertebrate rhodopsins could be functional expressed in other cell types and activate respective Gi/o signaling pathways in these hosts. We demonstrated this by co-expressing the Rat rhodopsin 4
(Rh) with GIRK channel subunits or Ca2+ channel in HEK293 cells. This demonstrated that in response to light, recombinant Rh could activate K+ channels and inhibit Ca2+
influx when expressed heterologously. We went further to express Rh in cultured
hippocampal neurons and showed membrane hyperpolarized within one second of light
exposure. This caused reduction in current induced firing rate of the neurons and
suggesting that Rh activates GIRK channels, localized somatodendritically (Mark and
Herlitze, 2000). In addition, analysis in autaptic neuron cultures (where neurons are
plated on microislands and are induced to synapse on themselves) revealed an increase in
paired pulse facilitation with light exposure (Li et al., 2005a). This form of short term
plasticity is governed by presynaptic mechanisms--i.e. Ca+ channels. The feasibility of the use of Rh as an optogenetic probe proved to be much greater than the invertebrate rhodopsins tested since Rh is able to activate endogenous Gi/o signaling.
85
The use of Rh to control intact circuits has been demonstrated by our group in
embryonic chicken spinal cord preparations. Electroporated Rh precisely inhibits the
spontaneous firing of the spinal cord motor neurons (Li et al., 2005a). The spinal cords
of chick embryos display rhythmic episodes of spontaneous bursting activity; the
frequency of this bursting is critical for motor axon pathfinding (Hanson et al., 2008).
Light application increased the intervals between bursting episodes which resulted in
~70% decrease of spontaneous motor unit activity. A decrease in asynchronous firing of
motor units between episodes was also observed in spinal cords. These effects can be
explained as consequences of membrane hyperpolarization. Interestingly, bilaterally
synchronized network bursting was stimulated upon cessation of light stimulus,
reminiscent of a rebound effect. This would most likely be due to relief of Na+ channel
inactivation (Li et al., 2005a).
The visual cycle of vertebrate and invertebrate rhodopins
To maintain the function of rhodopsin, the visual system must maintain the supply of the light sensitive chromophore, 11-cis-retinal. The series of transport and enzymatic reactions that govern retinoid homeostasis within the visual system is referred to as the visual cycle or retinoid cycle. In vertebrates, the retinal pigment epithelium (RPE), the tissue adjacent to photoreceptor cells, contains isomerases that convert vitamin A (all- trans-retinol) to 11-cis-retinal and supplies rhodopsin with photosensitive chromophore.
Exposure to light then converts 11-cis-retinal bound to lysine 296 to an all-trans
conformation. All-trans retinal must then be reconverted to 11-cis retinal for rhodopsin to regain activity.
86
Retinal regeneration after light stimulus starts when the Schiff base bond between
all-trans retinal and rhodopsin is hydrolyzed. The all-trans retinaldehyde is reduced to
all-trans retinol by retinol dehydrogenase (Saari et al., 1998). All-trans retinol then
leaves photoreceptor cells, crosses the interphotoreceptor matrix, apical RPE microvilli
and outer segment plasma membranes of photoreceptors. All-trans retinol then enters the
RPE cells. The intercellular transport of retinal is facilitated by binding to
interphotoreceptor or interstitial retinoid binding protein (IRBP), which is thought to
solubilize retinoids in the extracellular space and target the delivery of all-trans retinol to the RPE (Okajima et al., 1990; Pepperberg et al., 1993). Vitamin A (all-trans retinol) from the serum also enters the RPE, but enters from the basal surface (Gonzalez-
Fernandez, 2002). Once inside the RPE, all-trans retinol it binds to cellular retinol binding protein (CRBP) and is processed by an enzyme localized to the endoplasmic reticulum called lecithin retinol acyl transferase (LRAT). LRAT adds a fatty acid chain to all-trans retinol forming all-trans retinyl palmitate. This in turn becomes the substrate for a retinol isomerohydrolase, which catalyzes the formation of 11-cis retinol and palmitate (McBee et al., 2001; Rando, 1996). 11-cis retinol dehydrogenase oxidizes 11- cis retinol to 11-cis retinaldehyde which is subsequently released from the RPE. IRPB again facilitates extracellular transport by binding and targeting 11-cis retinaldehyde to photoreceptor cells where it reassociates with rhodopsin.
In contrast to vertebrate rhodopsin, invertebrate rhodopsins themselves possess the ability to regenerate 11-cis retinal through a photochemical reaction. In insects, the retinal chromophore remains bound to the rhodopsin apoprotein after light activation and the visual cycle does not require intercellular transport of retinoids. Photoconversion of
87
11-cis retinal to the all-trans conformation generates a thermally stable metarhodopsin
and a second photon (of higher wavelength) catalyzes reisomerization to 11-cis (Byk et
al., 1993). The relative stability of invertebrate metarhodopsin is not an intrinsic property,
but imparted by its interaction with arrestin (Kiselev and Subramaniam, 1994; Kiselev
and Subramaniam, 1996). Hence, the need for accessory retinoid binding proteins to
stabilize and traffic retinal is circumvented. The visual systems of cephalopods (which
include marine invertebrates such as squid, octopus, and cuttlefish) are similar to insects
in that all-trans retinal is reisomerized to 11-cis retinal via photochemical reaction, but differ from insects in that the Schiff base bond between all-trans retinal and rhodopsin is
hydrolyzed after activation. Unlike in insects, metarhodopsins of cephalopods are
thermally unstable and the spent chromophore binds soluble retinaldehyde binding
protein (RALBP) after metarhodopsin decay. RALBP facilitates delivery of all-trans
retinaldehyde to the retinochrome complex, present in the myeloid bodies of the
photoreceptor inner segments (Molina et al., 1992; Terakita et al., 1989). Absorption of a
second photon allows retinochrome to convert all-trans retinaldehyde back to the 11-cis
form, which then reassociates with RALBP and is recycled to the outer segments
(Terakita et al., 1989). Thus, in cephalopods, a rhodopsin-retinochrome conjugate system
maintains photoreceptive function in visual cells.
Retinoid processing in heterologous expression systems
In order to reconstitute function of invertebrate and vertebrate rhodopsins in non-
visual cell types, retinoid processing enzymes and binding proteins must be present to
sufficiently supply rhodopsin with photo-active substrate. Surprisingly, unlike the
88
photoreceptors of the vertebrate eye, it appears that many non-visual cell types possess the intrinsic capability to regenerate 11-cis retinal from all-trans-retinal. In human embryonic kidney cells (HEK293), activation of heterologously expressed rhodopsin can be achieved for over four hours once the cells are loaded with 11-cis retinal (Li et al.,
2005a). Furthermore, application of all-trans retinal, which presumable has to be reconverted to 11-cis retinal seems to be sufficient to achieve light responses in mammalian cells (Li et al., 2005a). Vertebrate rhodopsin expressed in HEK293S cells respond with early receptor currents, a conformation-associated charge shift, when exposed to light when supplied with 11-cis retinal, 9-cis retinal, 13-cis retinal and even all-trans-retinal or vitamin A (all-trans retinol) (Brueggemann and Sullivan, 2002).
However, responses could only be elicited after overnight incubations with all-trans
retinal or vitamin A. With respect to chromophore delivery, retinoids not only bind IRBP,
but also bind extracellular albumin in the interphotoreceptor matrix. This interaction may facilitate retinal trafficking within the visual context (Adler and Edwards, 2000). In the
heterologous setting, albumin may be able to take on the traditional role of binding proteins by solubilizing and successfully delivering retinal. In fact, fatty acid free bovine serum albumin has been used to successfully deliver retinal compounds to photoreceptors and HEK293 cells (Brueggemann and Sullivan, 2002; Li et al., 1999). For invertebrate rhodopsins expressed in non-photosensitive cells, a retinal compound must be initially supplied to confer functionality, but the photopigment should be able to regenerate active receptor itself since the chromophore remains bound to rhodopsin. Indeed, only a single
15 minute application of all-trans retinal is sufficient to generate receptor that is functional for several hours (Zemelman et al., 2002).
89
In the vertebrate eye, cultured hippocampal neurons, and the chick spinal cord, sufficient retinal compounds are available within neurons themselves to drive light activated currents even without an exogenous supply of chromophore (Li et al., 2005a).
This is consistent with the observation that in hippocampal neurons expressing the green algae Channelrhododopsin-2 (ChR2), no retinal compounds needed to be applied to activate the light-sensitive channel, though ChR2 requires all-trans retinal as chromophore (Boyden et al., 2005). This suggests that endogenous retinoids and/or what is provided in the culture medium are sufficient for rhodopsin loading, though it is possible that providing exogenous retinal may enhance light responses in non-visual cell types. Therefore, single application of 11-cis retinal, its analogs 9-cis or 13-cis, or even all-trans retinal is sufficient to regenerate the active compounds necessary for repetitive light activation of both vertebrate and invertebrate rhodopsins.
Melanopsin
Melanopsin is a light-sensitive pigment found on specialized retinal ganglionic cells that regulate circadian rhythms, pupillary light reflexes and other non-visual responses to light such as pineal melatonin synthesis (Melyan et al., 2005; Panda et al.,
2005). Melanopsin is a GPCR, but differs from other vertebrate opsin photopigments. In many respects, it actually resembles invertebrate rhodopsins, including its amino acid sequence and downstream signaling cascade via Gq and not the Gi/o pathway.
Additionally, melanopsin is a bistable photopigment, with intrinsic ability to convert all-
trans retinal to the active isomer 11-cis retinal (Panda et al., 2005).
90
The possibility of using melanopsin to impart light sensitivity by heterologous
expression has been explored by several groups. For example, melanopsin was co- expressed in Xenopus oocytes with arrestins and the transient receptor potential (TRP) channel, TRPC3, a mammalian homolog of the Drosophila phototransduction channels
TRP and TRPL. With light stimulation and application of all-trans retinal, melanopsin activates cation currents mediated by TRPC3 (Panda et al., 2005). Likewise, transiently
transfected recombinant melanopsin is able to activate TRPC3 channels in HEK293 cells
and Neuro-2a cells (mouse paraneuronal cell line) with light stimulus (Melyan et al.,
2005; Qiu et al., 2005). Heterologous expression in mammalian cells, did not require the
co-transfection of arrestins for melanopsin to transduce signal. However, to date, the
functional expression of melanopsin has not been demonstrated in neurons. It is unclear
whether melanopsin, like invertebrate rhodopsin requires the co-transfection of accessory
signaling molecules to function in neurons (Zemelman et al., 2002) or whether the
heterologous expression in neurons simply has not been attempted.
OptoXRs: chimeric GPCRs for the light-based control of specific GPCR intracellular
signaling pathways and behavior
Being GPCRs, vertebrate rhodopsins belong to the largest and most structurally
conserved family of signaling molecules. Rhodopsin belongs to the Class A, or
rhodopsin like, group of GPCRs and contains seven canonical transmembrane domains
(Fotiadis et al., 2006). The N-terminal domain faces the extracellular compartment, as do
the three extracellular loops. Rhodopsin also contains three intracellular loops and the
intracellular C-terminal domain. Structure-function studies of GPCRs and crystallization
91
of bovine rhodopsin suggest that there are conserved conformational mechanisms that occur during activation (Karnik et al., 2003; Palczewski et al., 2000). Based on this assertion, the intramolecular mechanisms of GPCR activation may be conserved, while molecular variations between related GPCR could confer differences in ligand affinity, G protein coupling specificity and interactions with other intracellular regulators.
Because GPCRs use common structural mechanisms to transduce signals, Dr.
Khorana and colleagues have generated functional chimeric GPCRs. They have shown that a mutant receptor in which the intracellular domains of vertebrate rhodopsin can be exchanged for those of the β2-adrenergic receptor to produce functional light-activated chimeric receptors (Kim et al., 2005). The G protein coupling specificity was converted from Gi/o to Gs, like the β2-adrenergic receptor. While the exchange of the 3rd intracellular loop (i3) was sufficient to induce adenylate cyclase activation, the efficiency was further increased by replacement of the i1-3 loops as well as the C-terminal domain.
This resulted in a light sensitive receptor that induced cAMP production approximately half as efficiently as agonist induced wild type β2-adrenergic receptor (Kim et al., 2005).
As an extension of this discovery, Dr. Deisseroth's group recreated the Rh/β2-
adrenergic described by the Khorana lab and also generated a chimeric Rh/α1-adrenergic
receptor. In response to light, Rh/β2 and Rh/α1 induced activation of Gs and Gq signaling
pathways in HEK293 cells, respectively (Airan et al., 2009). They call these chimeras
"OptoXRs." Expression of Rh/β2 and Rh/α1 in brain slices containing neurons from the nucleus accumbens induced increases in phosho-CREB suggesting that even downstream
signaling targets associated with the wild type adrenergic receptors can be stimulated by
light. Furthermore, light stimulus and positive expression could stimulate (with Opto-
92
α1AR) or inhibit (by Opto-β2AR) accumbens action potential firing. However, the most
surprising finding was the ability of Opto-α1AR to control reward-related behavior.
Optical fibers were implanted into the nucleus accumbens of mice virally transduced to
express Opto-α1AR in neurons. These mice were subjected to a place-preference test in
which light was applied when the animals entered a specific cage location. After one day of conditioning, the they were found to statistically "prefer" the cage location associated with light stimulus. In contrast, in vivo, stimulation using Opto-β2AR did not manifest
any changes in reward-related behavior (Airan et al., 2009). These results present the
possibility of broadly applying this technique to control other GPCR signaling cascades.
93
Research goals
The main goal of Part II of this dissertation is the generation of a light-activated G protein-coupled receptors that activate analogous downstream signaling to the wild type serotonin receptor, 5-HT1A. We hypothesize that since the intracellular protein domains
of 5-HT receptors contain the binding and interaction sites determining G protein
selectivity and subcellular localization, the molecular transfer of these critical protein
domains onto related GPCRs should also confer the ability to couple to specific
intracellular proteins necessary for targeting and function from the donor receptor. Since
vertebrate rhodopsin and 5-HT1A belong to the class A rhodopsin-like protein family, we
aimed to engineer chimeric rhodopsin/5-HT1A receptors for the control 5-HT1A signaling
in neurons and cells by light. We will attempt two different mutational strategies to
accomplish this goal.
The first is the exchange of the intracellular domains (three intracellular loops and
the C-terminal domain) of vertebrate rhodopsin with those of the 5-HT1A receptor. This
strategy is analogous to the approach taken previously by the Khorana and Deisseroth
laboratories to generate rhodopsin/adrenergic receptor chimeras (Airan et al., 2009; Kim
et al., 2005). Presumably, the result would be a GPCR that retains retinal binding and
light reactivity by maintaining the extracellular and transmembrane domains, but couple
to and traffic like the 5-HT1A receptor because the intracellular homology to 5-HT1A. The
second approach is to tag full length rhodopsin (without internal domain swaps) with
critical intracellular targeting domains from 5-HT1A. The rationale of this construct is
that since both rhodopsin and 5-HT1A are Gi/o coupled receptors, by targeting a Gi/o
94
coupled light sensitive receptor to subcellular sites where endogenous 5-HT1A is normally found, we will generate a functional substitute for 5-HT1A.
Experimentally, our main goals were to determine if the light-based activation of the chimeric receptors could functionally replace the endogenous, agonist-induced responses of 5-HT1A. We hoped to do this by cloning chimeric receptors and then functionally expressing the receptors in HEK293 cells to determine light sensitivity and the ability to activate Gi/o signaling. If the chimeras were functional in HEK cells, we planned to express them in cultured hippocampal neurons to further demonstrate functionality in a neuronal context. Then exogenous expression in neurons from 5-HT1A knockout mice could show that light-based activation of the chimera could substitute for ligand-activated responses of 5-HT1A receptors with its ability to functionally rescue KO phenotypes. Finally, expression of the chimeric receptor in dorsal raphe slices of KO mice could demonstrate that the light activated receptor could functionally replace endogenous 5-HT1A signaling in serotonergic neurons. Restoration of 5-HT1A-like responses by the chimeric receptor in cultured neurons and slices from KO mice would show functional redundancy.
95
CHAPTER 5
CONTROL OF 5-HT1A RECEPTOR SIGNALING BY
LIGHT-ACTIVATED G PROTEIN-COUPLED
RECEPTORS
96
Introduction
Serotonin (5-hydroxytryptamine; 5-HT) is an important regulator of various physiological functions and a critical modulator of disease. Most CNS structures receive serotonergic input arising mostly from neurons of the brainstem raphe nucleus. Of the 5-
HT receptors, most attention has been devoted to understanding the developmental and
behavioral effects associated with 5-HT1A (Barnes and Sharp, 1999; Patel and Zhou,
2005). The 5-HT1A receptor is of special interest because it is an important mediator of
many 5-HT functions, regulates the activity of 5-HT neurons as an autoreceptor, is
specifically implicated in the pathogenesis of anxiety and depression, and is thought to be
a physiologic target for selective serotonin reuptake inhibitor (SSRI) treatment (Gordon
and Hen, 2004; Hjorth et al., 2000; Lanfumey and Hamon, 2004). 5-HT1A receptors are
expressed widely throughout the CNS and periphery and are particularly enriched in limbic structures including the hippocampus and 5-HT system (Zhou et al., 1999). In
neurons, 5-HT1A is found on cell bodies and dendrites, allowing it to mediate 5-HT
effects on neuronal firing, both as auto- and as heteroreceptors (Patel and Zhou, 2005;
Riad et al., 2000; Sotelo et al., 1990). 5-HT1A activation leads to decreased firing via cell
membrane hyperpolarization; in 5-HT neurons, the consequence is inhibition of 5-HT
release (Stamford et al., 2000).
Despite being the subject of extensive study, there is still much ambiguity about
the normal in vivo function of 5-HT1A, dysfunction in the context of disease, and role in
therapeutic response. Traditional pharmacologic and genetic manipulations have been
enormously useful in defining in vivo 5-HT1A function. For example 5-HT1A receptor
agonists are anxiolytic (Gordon and Hen, 2004), 5-HT1A (-/-) mice reveal increased
97
anxiety (Ase et al., 2001; Heisler et al., 1998; Parks et al., 1998; Ramboz et al., 1998),
whereas transgenic expression of 5-HT1A receptors decrease anxiety (Kusserow et al.,
2004). Paradoxically, 5-HT1A mediated auto-regulation may underlie the 2-4 week
latency of SSRI efficacy. It has been suggested that substantial increases in 5HT release
would occur only after auto-receptor desensitization (Blier et al., 1987), and some report
that co-administration of 5-HT1A antagonists accelerates the onset of SSRI action (Artigas
et al., 1994; Hjorth, 1993; Hjorth et al., 2000). Constitutional 5-HT1A KO animals do not
exhibit SSRI-induced changes in behavior (Mayorga et al., 2001; Santarelli et al., 2003).
However, conditional knockdown of 5-HT1A auto-receptor expression in 5-HT neurons
imparts SSRI sensitivity to a genetic strain that is normally drug resistant (Richardson-
Jones et al., 2010). These studies clearly associate 5-HT1A with anxiety and response to
antidepressants, but also highlight the complex and poorly understood nature of in vivo 5-
HT1A signaling.
Drugs can be applied focally, but cannot be spatially contained or targeted to
specific cell populations; they also include side effects that can confound analysis.
Genetic mutations have the power to be targeted to specific cell types, but are
constitutionally lost or gained with KOs or transgenics, or imprecisely controlled with
conditional strategies. Neither approach can be controlled on a time scale of seconds.
We therefore aimed to develop an "optogenetic" tool to enable precise analysis of in vivo
5-HT1A function by using light activated G protein-coupled receptors (GPCRs), like those found in the visual system. We previously demonstrated that the GPCR, vertebrate rhodopsin (Rh), can be functionally expressed in non-visual cell types to activate
downstream targets of Gi/o signaling (Li et al., 2005a). Furthermore, Rh can be
98
exogenously expressed in neurons in primary culture and in intact animals to inhibit
neuronal and neural network excitability (Li et al., 2005a). Based on these findings we
aimed to tailor the properties of Rh to manipulate other GPCR signaling pathways,
namely, the 5-HT1A. Here we describe the development of a chimeric light-sensitive
GPCR that mimics the intracellular targeting and functional Gi/o-linked signaling of wild type 5-HT1A. This receptor, which we call Rh-CT5-HT1A, is able to functionally substitute for endogenous 5-HT1A receptors by exploiting the intracellular trafficking mechanisms used by the endogenous receptors. Rh-CT5-HT1A distributes intra-neuronally to cell
membrane sites normally occupied by 5-HT1A, which then allows Rh-CT5-HT1A to induce
activation of the same downstream Gi/o signaling targets with light stimulus. We reason
that this chimera can then be used as a proxy to modulate 5-HT1A-like activity in intact, behaving animals using non-invasive techniques. Hence, this tool could prove to be a significant advance, by allowing the precise characterization of in vivo 5-HT1A function.
99
Materials and Methods
Generation of Plasmid Constructs for Transfection and Pseudovirion Production
Rat Rh (RO4) and human 5-HT1A cDNA (GenBank accession nos. Z46957 and
AF498978) clones were tagged C-terminally with mCherry immediately following the
last coding codon using a 2-step fusion PCR. Distal primers for Rh-mCherry were 5´-
ATCGCTCGAGATGAACGGC ACAGAGGGC-3´ and 5´-GCTGATTATGATCT
AGAGTCGCG-3'; distal primers for 5-HT1A-mCherry were 5´-
ATCGCTCGAGATGGATGTG CTCAGCCCTG-3' and 5´-GCTGATTATGATCT
AGAGTCGCG-3'. Primers for the fusion site were 5´-
AGCCAGGTGGCTCCAGCCATGGTG AGCAAGGGCGAG-3´ and 5´-CTCGCCCTT
GCTCACCATGGCTGGAGCCACCTGGCT-3´ for RO4-mCherry and 5'-
ATTAAGTGTAAGTTC TGCCGCCAGATGGTGAGCAAGGGCGAG-3' and 5'-
CTCGCCCTTGCTCACCATCTGGCGGC AGAACTTACACTTAAT-3' for 5-HT1A- mCherry. For Rh-CT5-HT1A, the C-terminal domain of Human 5-HT1A was appended immediately following the last coding nucleotide of Rh-mCherry by fusion PCR using the 5' distal primer for Rh-mCherry, the 3' distal primer for 5-HT1A-mCherry and the fusion primers 5'-GCATGGACGAGCTGTACAAGAACAAGGAC
TTTCAAAACGCG-3' and 5'-CGCGTTTTGAAA
GTCCTTGTTCTTGTACAGCTCGTCCATGC-3'. These fragments were PCR- amplified using the respective distal primers and cloned into the XhoI and NotI sites of pEGFP-N1 (Clontech) to generate HEK cell expression clones. Human 5-HT1A cDNA was purchased from the Missouri S&T cDNA Resource Center (Rolla, MO).
100
To construct Sindbis virus expression vectors, SinRep(nsP2S726)dSP-EGFP (Li et
al., 2005a) was modified using the gateway vector conversion system (Invitrogen).
Briefly, the RfA cassette was cloned into the PmlI site of SinRep(nsP2S726)dSP-EGFP to generate a gateway destination vector. Entry clones were generated by cloning of genes of interest into pENTR/D-TOPO or pCR8/GW/TOPO according to manufacturer's protocol (Invitrogen). LR recombination was performed to generate final Sindbis expression clones. Lentivirus expression constructs were made by LR recombination of each entry clone together with pENTR5'/CMVp into pLenti6.4/R4R2/V5-Dest
(Invitrogen).
Cell Culture, Virus Production and Infection
Cell culture and maintenance of Human embryonic kidney (HEK) 293 cells
(tsA201 cells) were performed as described previously (Wittemann et al., 2000). Cells were transfected with 2 μg of each GPCR DNA and 1 μg of each GIRK channel subunit
DNA with Lipofectamine 2000 (Invitrogen) and incubated for 18-24 h prior to recordings or fixation for immunocytochemistry.
Sindbis pseudovirions were generated as previously described (Li et al., 2005a).
Lentiviral particles were made by cotransfection of each Lentivirus expression vector together with pLP1, pLP2 and pLP/VSVG helper plasmids into HEK 293T/17 cells
(ATCC# CRL-11268) according to Invitrogen protocols. Both Sindbis and Lentiviral particles were concentrated by ultracentrifugation at 160,000 x g for 90 min through a
20% sucrose cushion and resuspended in Hanks' Balanced Salt Solution. Viral titer was greater than 1x108 units per ml and stocked at -80°C.
101
Continental culture of hippocampal neurons from P0-P3 rats and mice were
performed by a modified Banker sandwich method as described (Bekkers and Stevens,
1991; Xie et al., 2007). The generation of 5HT1A KO mice (Heisler et al., 1998; Parsons
et al., 2001) and genotyping methods (Scott-McKean et al., 2008) have been previously
described. WT mice (C57BL/6J) were obtained from Jackson Laboratories (Bar Harbor,
ME). Handling and care of mice followed federal guidelines and experimental methods
were approved by the Case Western Reserve University Institutional Animal Care and
Use Committee. For neuronal infection, 0.5-5 µls of thawed Sindbis virus suspension
was added to cultured hippocampal neurons (9-14 DIV) on coverslips in 24-well plates.
GFP expression was detected after 10 h and reached maximal expression after 24 h.
Lentiviral injections into the dorsal raphe nucleus
Lentivirus expressing Rh-CT5-HT1A was injected into the dorsal raphe nucleus
(DRN) of wild type (C57Bl/6J), ePet::YFP (Scott et al., 2005) or 5-HT1A(-/-) mice (Heisler et al., 1998). 3 week old mice were anesthetized with 1-2% isoflurane in air delivered from a precision vaporizer (WPI) and mounted onto a stereotactic frame (Narishige). A sagittal incision along the midline was made to expose the cranium and a burr hole was drilled 4.1 mm caudal from the bregma. The tip of a micropipette attached to a 30 ml syringe was lowered into the dorsal raphe nucleus and 3-4 µl of the virus was injected.
Mice were housed for 7-10 days before performing immunohistochemistry or electrophysiological experiments.
Immunofluorescence, Image Acquisition and Data Analysis
102
tsA201 cells were transfected with indicated DNAs using Lipofectamine 2000 and
hippocampal neurons (8-10 DIV) were transfected with using Sindbis viral stocks. HEK
cells and neurons were fixed with 4% paraformaldehyde for 10 min and permeabilized
with 0.1% Triton X-100 in PBS 18 h and 12 h post-transfection, respectively. Anti- dsRed (Clontech; 1:300) was used to label mCherry tagged receptors and anti-5-HT1A
(Millipore; 1:500) was used to stain endogenous receptor. Anti-MAP-2 (Sigma; 1:500) and anti-Tau-1 (Millipore:1:200) were used to label dendritic and axonal processes. Cells
were blocked with 10% normal goat serum and 3% BSA and incubated with primary
antibody overnight at 4°C. After extensive washes, they were incubated with Alexa 405-
and Alexa 546- conjugated secondary antibodies (Molecular Probes) for 30 min at room
temperature. Cells were mounted in Prolong Gold antifade medium (Molecular Probes).
Images were acquired with a Zeiss LSM 510 confocal microscope using 20x and 40x
water objectives and analyzed by using VOLOCITY (Improvision, Lexington, MA) and
Zeiss LSM 5 software (Rel. 3.2). Z-stack images were acquired to image the entire cell
and displayed as a projected image or single slice through the center of cell where
indicated. For quantification of relative fluorescence intensity, imaging parameters were
adjusted so that pixel intensity within neurites did not saturate. The line profile function
in the LSM 5 software was used to trace the longest dendrite of each neuron analyzed.
Dendrites were identified by both MAP-2 (dendritic marker) and GFP (positive infection)
fluorescence. Fluorescence intensity was normalized to maximal intensity of each
dendrite. For quantification of fluorescence along dendrite, piecewise linear interpolation
was performed of each plot to normalize the line profile distance to 1000 values between
103
0.0 and 1.0. Interpolated data were then grouped and plotted as the mean ± S.E.M. at
each normalized point.
For immunohistology, adult mice were deeply anesthetized using Avertin (0.5g
tribromoethanol/39.5ml H20, 0.02 ml/g body weight) before transcardial perfusion with
4% paraformaldehyde (PFA) in 0.1M PBS for 20 min. The brain was then removed and
fixed in PFA for another 2 h at room temperature (RT) followed by cryoprotection in
30% sucrose (w/v) overnight at 4°C. Tissue sections (16-20 µm) were prepared on a
freezing microtome or cryostat and mounted on Superfrost Plus Microscope Slides
(Fisher Scientific, Pittsburgh, PA). Fluorescent immunohistochemistry was performed
as described (Lerch-Haner et al., 2008). Tissue sections were immunolabeled with rabbit
anti-GFP (1:1,000, Invitrogen, Carlsbad, CA) and then FITC conjugated secondary
antibody (1:200, Jackson ImmunoResearch, West Grove, PA). Fluorescent images were
collected using a SPOT RT color digital camera (Diagnostic Instruments, Sterling
Heights, MI) attached to an Olympus Optical BX51 microscope (Center Valley, PA).
Electrophysiology and Data Analysis
For GIRK channel recordings, Human GIRK channel subunits (KCNJ3/5) and
light sensitive GPCRs or 5-HT1A receptor were coexpressed in tsA201 cells. GIRK
subunit DNA was purchased from Genecopoeia (Rockville, MD). Cells were cultured
and recorded in dark room conditions (red light only) following transfection. GIRK- mediated K+ currents were measured and analyzed as described previously (Li et al.,
2005b). Absolute GIRK current was determined by brief application of a low K+ solution
to abrogate GIRK current: 138 NaCl, 2 KCl, 2 CaCl2, 1 MgCl2, 10 HEPES-NaOH, pH
104
7.3 (KOH). The difference in current elicited with high K+ (external solution) and low
K+ solutions was determined to be the absolute GIRK current. The external solution was as follows (mM): 20 NaCl, 120 KCl, 2 CaCl2, 1 MgCl2, 10 HEPES-NaOH, pH 7.3
(KOH). Patch pipettes (2-5 MΩ) were filled with internal solution (mM): 100 potassium aspartate, 40 KCl, 5 MgATP, 10 HEPES-KOH, 5 NaCl, 2 EGTA, 2 MgCl2, 0.01 GTP, pH 7.3 (KOH). Cells were incubated in external solution containing 1 µM 9-cis retinal
(Sigma) for 20 min prior to light stimulation. Cells were visualized using transilluminated red light (590 nm filter) during experimental manipulations. Guanosine
5'(γ-thio)triphosphate (GTPγS) was added to the internal solution at a final concentration of 0.6 mM where indicated. Solutions containing agonist or low potassium were applied directly onto the recorded cells using a fast-flow perfusion system (ALA Scientific,
Farmingdale, NY).
Cultured hippocampal neurons were used for recordings 10-14 day in vitro, 14-20 h after Sindbis virus infection. Extracellular recording solution contained (mM): 125
NaCl, 2 KCl, 10 Hepes, 30 glucose, 3 mM CaCl2, and 1 MgCl2, pH 7.3 (NaOH); internal solution contained (mM): 97 potassium gluconate, 10 Hepes, 1 potassium-EGTA, 4 Mg-
ATP, and 0.4 mM Na-GTP pH 7.3 (KOH). Synaptic activity was silenced by adding 10
µM 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, Tocris) and 10 µM SR 95531 hydrobromide (Gabazine, Tocris). Cells were perfused with 1 µM 9-cis retinal (Sigma) for 2 min before light stimulation. 1 µM 8OH-DPAT (Calbiochem) and 50 µM baclofen
(Sigma) were used in experiments where indicated.
Whole cell patch clamp recordings of cultured neurons and tsA201(Hamill et al.,
1981) were performed with an EPC9 amplifier (HEKA). Currents were digitized at 10
105
kHz and filtered with the internal 10-kHz three-pole Bessel filter (filter 1) in series with a
2.9-kHz 4-pole Bessel filter (filter 2) of the EPC9 amplifier. Series resistances were
partially compensated between 70 and 90%.
Brain slice recordings
Coronal slices including dorsal raphe (250 μm thick) were cut from brainstems of
the mice 8-12 days after Lentivirus injection. Mice were anesthetized with isoflurane and
decapitated. The removed brainstem was cooled and sliced in ice-cold solution
containing (in mM) 87 NaCl, 75 Sucrose 2.5 KCl, 0.5 CaCl2, 7 MgCl2, 1.25 NaH2PO4, 25
NaHCO3 and 20 glucose bubbled with 95% O2 and 5% CO2 using with a vibratome
(VT1000S, Leica). Slices were stored for at least 1 hour at room temperature in a recording artificial cerebrospinal fluid containing (in mM) 124 NaCl, 3 KCl, 2.5 CaCl2,
1.2 MgSO4, 1.23 NaH2PO4, 26 NaHCO3 and 10 glucose bubbled with 95% O2 and 5%
CO2. Fluorescent mCherry positive cells were visually identified under an upright
microscope (DMLFSA, Leica) equipped with a monochromator system (Polychrome IV,
TILL Photonics) flashing 585 nm excitation light. Whole-cell recordings were made at room temperature in the dark except for using infrared light to target the cell. Slices were pre-incubated at least 20 min and continuously perfused with the external solution including 25 μM 9-cis retinal, 0.025% (±)-α-tocopherol (Sigma), 0.2% essentially fatty acid free albumin from bovine serum (Sigma) and 0.1% dimethyl sulfoxide. Patch pipettes (2-4 MΩ) were filled with an internal solution with the following composition
(in mM) 140 K-methylsulfate, 4 NaCl, 10 HEPES, 0.2 EGTA, 4 Mg-ATP, 0.3 Na-GTP
and 10 Tris-phosphocreatine, pH 7.3 (KOH). Membrane currents and voltages were
106
recorded with an EPC10/2 amplifier (HEKA). The signals were filtered at 3 kHz and
digitized at 50 kHz. The PatchMaster software (HEKA) was used for the controls of
voltage and data acquisition, and off-line analysis was made with Igor Pro 6.0 software
(Wavemetrics).
Statistical significance throughout the experiments was tested with ANOVA
(Tukey's post-hoc test) using Prism (GraphPad) or Igor 6.0 software (Wavemetrics).
Standard errors are mean ± S.E.M.
107
Results
Cloning of Rh-CT5-HT1A and optimization of the light activation paradigm.
Vertebrate rhodopsin (Rh) and 5-HT1A are Gi/o-linked GPCRs belonging to the
Class A (rhodopsin-like) family of seven transmembrane domain receptors. We hypothesized that we could functionally replace native 5-HT1A with a Gi/o-coupled light- sensitive receptor by inducing its subcellular targeting to mimic that of wild type 5-HT1A.
Since the determinants of subcellular targeting are presumably contained within the
intracellular domains of GPCRs, we reasoned that a Gi/o-coupled light-activated receptor
could adopt the intracellular targeting properties of 5-HT1A if key trafficking domains of
5-HT1A were added. We did this by tagging Rh with the C-terminal domain (CT) of 5-
HT1A receptor, which has been shown to be a critical determinant for correct
somatodendritic trafficking of 5-HT1A via its interaction with the putative ER/Golgi
trafficking protein,Yif1B (Carrel et al., 2008). This receptor, which we call Rh-CT5-HT1A,
consists of Rh tagged C-terminally with mCherry and then with the CT of 5-HT1A (Fig.
7A). To determine if this modified receptor retained Gi/o-linked GPCR activity, it was co-expressed with Human GIRK 1 and 4 subunits in HEK293 cells.
Accurate functional comparison of chimeric receptors with Rh and other GPCRs required modification of the fluorescent tag, retinal loading and recording conditions to improve assay consistency. GFP has been previously used as a C-terminal tag to track functional expression in transgenic animals (Jin et al., 2003; Moritz et al., 2001; Perkins et al., 2004). However, when we expressed a similar construct, Rh-EGFP, in HEK293 cells, the amplitude of maximal GIRK channel activation we observed was at best only
~50% of that of untagged or mCherry-tagged versions of Rh (data not shown).
108
Furthermore, the consistency of responses to light stimulus of GFP positive cells was
relatively low, 44.4% (12/27). Rh is maximally excited at 485 nm which coincides
almost exactly with the excitation wavelength of EGFP (488 nm). Thus, use of EGFP as
a marker for positive transfection could cause inadvertent receptor activation since the
rhodopsin apoprotein itself (even in the absence of retinal) exhibits weak activity (Melia
et al., 1997). mCherry is an improved fluorescent tag because it has an
excitation/emission profile of 587nm/610nm, which lies outside of the absorption
spectrum of Rh (Shaner et al., 2004).
Phototransduction by Rh is initiated by the isomerization of the photosensitive
pigment, 11-cis retinal, by light. In the visual system, spent substrate (all-trans retinal) is
recycled by a series of transport and enzymatic reactions (Lamb and Pugh, 2004). HEK
cells possess the intrinsic capability to regenerate 11-cis retinal from all-trans retinal or other analogs such as 9-cis or 13-cis retinal, but require an exogenous source of retinal
(Brueggemann and Sullivan, 2002). Another source of variability was the retinal loading conditions of Rh and variants, which was confounded by the variability of serum used as a culture media supplement. Fetal bovine sera (FBS) contain retinal compounds as evidenced by the ability to activate transfected Rh in HEK293 cells cultured in media made with some, but not all lots of FBS. This raised the possibility that ambient light could inadvertently activate the light sensitive GPCRs. This in turn could lead to a decrease in receptor activity and/or desensitization, potentially confounding the experiments. Considering these complications, cells were kept in the dark following transfection and during experimental procedures. Furthermore, a 20 min preincubation of
1 µM 9 cis-retinal prior to recordings was used to yield the most consistent results,
109
regardless of culture media composition. Over 90% (19/21) of mCherry positive cells responded to light stimulus under these optimized conditions.
In vertebrate rod and cone cells, bleached vertebrate rhodopsin is able to transduce signal and the pigment may remain in a steady state of activation even after light stimulation is eliminated (Cornwall and Fain, 1994; Cornwall et al., 1995; Fain et al., 1996). Thus, we tested the possibility that light activated receptors were active in heterologous expression systems even in the absence of light stimulus. This phenomenon could limit the extent of GIRK current modulation observed and could lead to constitutive, basal increases in Gi/o activation. More importantly, since the ultimate goal
is to exogenously express Rh-CT5-HT1A in other cell types, this would limit the utility of the light sensitive receptor since merely expressing it would affect baseline properties
without light application. HEK293 cells transfected with GIRK1/4 subunits alone (Fig.
1A) or co-transfected with Rh (Fig. 1B) were analyzed with GTPγS (a non-hydrolyzable
GTP analog) in the intracellular recording solution. GTPγS caused constitutive G protein activation and gradually led to maximal GIRK current induction. The absolute GIRK current was assessed by a short application of low K+ (2 mM), eliminating the inward K+
current. The GIRK current was then calculated as the difference between current
immediately before (high K+) and during the low K+ treatment. The GIRK currents
induced by GTPγS were not significantly different for HEK cells transfected with Rh and
GIRK [411 ± 65 pA (n = 8)] versus GIRK subunits alone [376 ± 47 pA (n = 9)] (Fig.
1C), indicating that there was no appreciable activation of GIRK by Rh without light
stimulus. Furthermore, the maximal GIRK current revealed by GTPγS was comparable
in cells transfected with GIRK [772 ± 182 pA (n = 9)] or Rh with GIRK subunits [793 ±
110
79 pA (n = 8)] (Fig. 1D), suggesting that Rh co-expression did not interfere with GIRK
expression level or targeting.
Expression pattern and function of Rh-CT5-HT1A resembles fluorescently tagged 5-HT1A
in HEK cells.
The determinants of G protein specificity and subcellular targeting are
presumably contained within the intracellular domains of GPCRs. Based on this
assertion, two different chimeric receptors were generated. The first consisted of Rh
tagged C-terminally with mCherry and then the CT domain of 5-HT1A (Rh-CT5-HT1A).
The CT domain of 5-HT1A has been shown to be necessary for its dendritic targeting via
interaction with a putative ER/Golgi trafficking protein, Yif1B (Carrel et al., 2008;
Jolimay et al., 2000). We reasoned that transferring a critical targeting domain would
also transfer the ability to traffic in an analogous fashion to 5-HT1A. The second
construct was a mutant Rh receptor in which the intracellular domains were exchanged
for those of the 5-HT1A receptor. The rationale for this GPCR was that extracellular and transmembrane domains of Rh were retained, thus preserving responsiveness to light, but
the intracellular domains of 5-HT1A would induce subcellular targeting and G protein coupling like 5-HT1A. mCherry tagged versions of Rh, 5-HT1A, and Rh-CT5-HT1A were
cloned into mammalian expression vectors (Fig. 5A). Protein alignments of these three
GPCRs are shown in Fig. 2. For the chimera containing 5-HT1A intracellular loop and
CT, the exact Rh residues retained and domain borders were analogous to the Rh/β2-
adrenergic and Rh4/α1-adrenergic receptors generated by the Khorana and Deisseroth
groups (Airan et al., 2009; Kim et al., 2005). A high degree of similarity is observed
111
between the transmembrane regions of Rh, 5-HT1A, β2-adrenergic and α1-adrenergic receptors (Fig. 3). However, this receptor revealed uncharacteristic kinetics, taking several minutes to fully activate and was constitutively active once light was applied (Fig.
4). For these recordings, cells were perfused with a low K+ solution every two minutes
to monitor real K+ currents and rule out artifactual drift (which could resemble increased
inward current) that may occur with long patch clamp recordings (Fig. 4C). Because
GIRK current induction coincided with light application (Fig. 4B) and the total current
induced was similar to other GPCRs tested, we concluded that the receptor was
"functional." However, because of the unusually slow kinetics and constitutive nature of
activation, we did not perform a more precise analysis of this chimeric receptor, but
concentrated on the characterization of Rh-CT5-HT1A.
When co-transfected into HEK293 cells, exogenously expressed Rh-mCherry,
Rh-CT5-HT1A, and 5-HT1A-mCherry targeted efficiently to the cell membrane and
colocalized with GIRK channel subunits (Fig. 5B). Functionally, Rh-mCherry and Rh-
CT5-HT1A were able to activate GIRK current when exposed to light at 485 nm (Figs. 5C
& 5D). The extent of GIRK activation was not significantly different from untagged Rh
and was similar to responses induced in tagged and untagged 5-HT1A receptors (Figs. 5E
& 5F) by the selective 5-HT1A agonist, 8OH-DPAT (Fig. 5G). It is important to note that the GIRK current induction for all GPCRs tested is similar to that induced by GTPγS
(Fig. 1B), suggesting that both agonist and light application caused near maximal induction of GIRK for 5-HT1A and light sensitive receptors, respectively. The time
constants for onset of GIRK channel activation and deactivation were also similar
between Rh, Rh-CT5-HT1A and 5-HT1A (τon ≈ 2-10 s, τoff ≈ 30-50 s, Fig. 5H), though
112
activation of 5-HT1A and 5-HT1A-mCherry were significantly faster (τon = 1-2 s) than the
light activated receptors (τon ≈ 9-10 s). Activation time courses for light activated
receptors were not significantly different, nor did activation of tagged versus untagged 5-
HT1A differ. The kinetics of inactivation for all GPCRs tested were not significantly
different, suggesting that addition of mCherry and CT tags to Rh does not interfere with
function.
When recording light-induced currents from HEK cells, we noticed an interesting
phenomenon with successive recordings in a single cell. In this protocol, cells co-
expressing Rh, Rh-mCherry or Rh-CT5-HT1A with GIRK 1/4 subunits were stimulated
with 10 s pulse of light as described above. After allowing the cell to recover (75 s)
another 10 s light pulse was applied and allowed to recover again (Fig. 6). Second and third light applications enhance GIRK current to a significantly lower degree (~50% -
60% for second and < 50% for the third) when compared to the first light exposure (Fig.
6B). This phenomenon was observed for both untagged and tagged Rh as well as Rh-
CT5-HT1A (Fig. 6B). The most likely and straightforward explanation of this phenomenon
is that after a 10 s light pulse the retinal substrate has been depleted from Rh and cannot
be restored to naive loading conditions in the given recovery time frame. Another, more
interesting cause could be that retinal levels are not limiting, but rather after stimulation,
GPCR desensitizes and/or GIRK channels are trafficked away from the membrane.
Regardless, the decrement in successive responses highlights the importance of
preventing premature light exposure to cells expressing light-sensitive GPCRs.
113
Subcellular targeting of Rh-CT5-HT1A resembles that of 5-HT1A.
Exogenous expression of Rh targets to both somatodendritic and presynaptic sites when expressed in rat hippocampal neurons (Li et al., 2005a). We confirmed this by immunolabeling neurons infected with Sindbis virus driving the expression of Rh- mCherry and also EGFP under the control of a second subgenomic promoter. GFP was expressed throughout the entire cell and proved to be a valuable tool for identifying and matching axons, dendrites and soma of infected neurons. Rh-mCherry (Fig. 7A)
colocalized with the dendritic marker, MAP-2 (Fig. 7C), but was also expressed in processes that were presumably axons because they were GFP positive (Fig. 7B), but lacked MAP-2 labeling (Fig. 7A-D, n = 10). To confirm this, neurons were stained with anti-Tau-1 antibody (axonal marker) and we observed Rh-mCherry fluorescence that coincided with processes labeled by Tau-1 (Fig. 7E-H, n = 10). This indicated that Rh targeted both to axons and dendrites confirming our previous findings (Li et al., 2005a).
In contrast, virally expressed 5-HT1A-mCherry localized somatodendritically in neurons
(Fig. 7M-P, n = 8), which is consistent with the in vivo distribution of 5-HT1A observed in
serotonergic and hippocampal neurons (Kia et al., 1996; Patel and Zhou, 2005; Riad et al., 2000; Sotelo et al., 1990). Axons (Tau-1 and GFP positive) in these neurons lacked
5-HT1A-mCherry labeling indicated by dsRed staining. Likewise, Rh-CT5-HT1A showed
an analogous intracellular trafficking pattern (Fig. 7I-L, n = 12). Neurons infected with
Rh-CT5-HT1A virus showed colocalization of Tau-1 and GFP fluorescence in axonal
processes, but not dsRed and Tau-1. No dsRed+/GFP+/MAP-2 negative processes were
seen in neurons infected with 5-HT1A or Rh-CT5-HT1A viruses (data not shown). This
indicated that Rh-CT5-HT1A and 5-HT1A were absent from the axonal processes of
114
positively infected neurons (Tau-1+/GFP+) at described expression conditions. Axonal
Rh-CT5-HT1A was only observed when the receptor was expressed at very high levels by
using 2-10 fold higher titer virus and allowing Sindbis infection to occur for greater than
24 hours (data not shown). At these conditions, toxicity effects and cell death were
observed most likely due to inhibition of host protein synthesis by excessive virally
driven expression (Kim et al., 2004). Axonal targeting and toxicity were similarly
observed with high expression of wild type 5-HT1A. Taken together, the data show that
the CT of 5-HT1A is sufficient to promote somatodendritic trafficking away from axons in
an analogous manner to wild type 5-HT1A.
Another similarity between 5-HT1A-mCherry and Rh-CT5-HT1A was their efficient
targeting to the distal ends of dendrites. In comparison to Rh-mCherry (Fig. 8A), 5-
HT1A-mCherry and Rh-CT5-HT1A fluorescence was observed much further away from the
soma (Fig. 8B-C). To quantify dendritic fluorescence distributions, neurons infected with
Sindbis virus were stained with anti-dsRed to enhance the mCherry signal and minimize potential bleaching artifacts introduced by tracking mCherry fluorescence alone.
Neurons were also stained with anti-MAP2 antibody to label dendrites. The longest dendrite of infected cells (MAP2+/GFP+) were analyzed in a similar way as described previously (Carrel et al., 2008). Normalized fluorescence was quantified with respect to both absolute distance from the soma (Fig. 8E) and relative distance along the dendrite
(Fig. 8F). 5-HT1A-mCherry and Rh-CT5-HT1A were present even at the distal ends of all
dendrites. In contrast, Rh-mCherry fluorescence decayed significantly faster along the
length of the dendrite, with most of the appreciable fluorescence restricted to the
proximal half of the dendrite. The difference in fluorescence distribution was not due to
115
effects on neuronal morphology since the lengths of the dendrites examined were not statistically different between groups (Fig. 8D). Thus, CT of 5-HT1A promotes distal
targeting and is sufficient to drive the trafficking of Rh-CT5-HT1A analogously to wild type
5-HT1A receptor.
Rh-CT5-HT1A is functional in cultured hippocampal neurons and competitively inhibits
endogenous 5-HT1A receptor.
A consequence of Gi/o-linked signaling in neurons is the activation of GIRK
channels, which are predominantly expressed in dendrites (Mark and Herlitze, 2000).
GIRK channel activation causes an efflux of K+ resulting in hyperpolarization.
Therefore, we tested the ability of Rh-CT5-HT1A to activate Gi/o signaling and induce hyperpolarization in a neuronal context. mCherry tagged Rh (Fig. 9A) as well as the Rh-
CT5-HT1A chimera (Fig. 9B) caused a 8-9 mV membrane hyperpolarization (postsynaptic
effect) in response to a 1 s light pulse. Hyperpolarization was sustained for the duration
of longer (10 s) light stimulus protocols and cells showed rapid reversal of membrane
voltage change after light was turned off (data not shown). Uninfected hippocampal
neurons were also assessed for their ability to respond to baclofen, a GABAB agonist which served as a positive control for Gi/o activation, and the selective 5-HT1A agonist,
8OH-DPAT. For quantification of biophysical properties, a 1 s light or agonist
application was used, which was long enough to induce maximal activation of GPCR and
induce hyperpolarization but short enough so that GPCR desensitization was not
observed. The resulting changes in membrane voltage for Rh-mCherry and Rh-CT5-HT1A
stimulated by light were similar to neuronal responses with activation of endogenous
116
GABAB or 5-HT1A receptors (Fig. 9C). The time constants for hyperpolarization and recovery by GPCR activation in neurons were much faster than in HEK293 cells (Fig. 9D versus 3H). This is most likely due to the effect of proteins endogenous to neurons, such as RGS proteins, which potentiate the GTPase activity of G proteins (Mark and Herlitze,
2000).
Rh-CT5-HT1A contains protein sequence for the interaction with the trafficking
machinery normally used by endogenous 5-HT1A receptors. This allows analogous
targeting as 5-HT1A, but also induces a dominant negative effect because Rh-CT5-HT1A
could presumably compete to interact with the same intracellular trafficking proteins
(Fig. 10). Hyperpolarization induced by 8OH-DPAT in hippocampal neurons expressing
Rh-CT5-HT1A was decreased to 47% ± 9 (n = 10) of the non-transfected neuron responses
recorded in parallel (Fig. 10). Rh-mCherry expression did not affect 8OH-DPAT responses indicating that neither viral infection nor exogenous expression of GPCRs affect endogenous 5-HT1A responses. This effect is consistent with the experiments
showing that co-expression of the CT itself reduces 5-HT1A in distal dendrites in cultured
neurons (Carrel et al., 2008). Application of baclofen resulted in comparable
hyperpolarization for uninfected versus Sindbis virus treated neurons, ruling out the
possibility that virus application interrupted targeting and expression for all endogenous
GPCRs.
117
Rh-CT5-HT1A compensates for the loss of 5-HT1A signaling in cultured hippocampal
neurons of 5-HT1A null mice.
To demonstrate that Rh-CT5-HT1A could functionally substitute for 5-HT1A
receptors we sought to determine if Rh-CT5-HT1A could functionally "rescue" 5-HT1A
signaling in neurons from 5-HT1A KO mice. As expected, 5-HT1A immunostaining was
absent from neurons from 5-HT1A null mice (Heisler et al., 1998), but hippocampal neurons of wild type mouse showed robust staining throughout dendrites (Fig. 11A).
Functionally, the application of 1 µM 8OH-DPAT (5-HT1A agonist) onto neurons of KO
mice failed to elicit a hyperpolarization response, even with longer agonist applications
(10 s) (Fig. 11E). WT mouse neurons hyperpolarized when exposed to 1 µM 8OH-
DPAT with a comparable response to what was seen in hippocampal neurons cultured
from wild type rats (Fig. 11C and 11H versus Fig. 9C). The response to baclofen
remained intact in KO neurons suggesting that the mutation is specific to 5-HT1A and does not affect Gi/o signaling broadly (Fig. 11D and 11H). 5-HT1A-mCherry and Rh-CT5-
HT1A expressed by Sindbis virus vectors localized to the dendrites of KO neurons (Fig.
11A). The hyperpolarization defect in KO neurons was rescued by exogenous expression
of both 5-HT1A-mCherry (with agonist application, Fig. 11F) and Rh-CT5-HT1A (with
light, Fig. 11G). The loss of function phenotype was completely compensated for since
activation of exogenously expressed 5-HT1A and Rh-CT5-HT1A was indistinguishable from
wild type mouse neuron response to 8OH-DPAT.
118
Rh-CT5-HT1A functionally substitutes for 5-HT1A signaling in dorsal raphe nucleus
neurons in brain slices from 5-HT1A null mice.
We next wanted to demonstrate if Rh-CT5-HT1A was capable of modulating
serotonergic neurons of the dorsal raphe in a analogous manner to endogenous 5-HT1A.
Again, these neurons are important players in the pathogenesis of anxiety and depression.
Lentivirus expressing Rh-CT5-HT1A was stereotactically injected into the dorsal raphe
nucleus (DRN) of ePet::YFP (Scott et al., 2005) or 5-HT1A (-/-) mice (Heisler et al.,
1998). As indicated in Figure 8A, Rh-CT5-HT1A was expressed in 5-HT neurons (labeled
with YFP in ePet::YPF mice) and revealed a punctate distribution most prominently in
the soma. We next determined if Rh-CT5-HT1A could functionally rescue the phenotype in
DRN neurons of 5-HT1A KO mice. Similar to what was observed in cultured
hippocampal neurons, application of the 5-HT1A agonist, 8OH-DPAT (1 µM), onto
brainstem slices failed to elicit a hyperpolarization response in DRN neurons of KO mice
(data not shown). This defect can be rescued by expression of Rh-CT5-HT1A and
subsequent activation by light (Figure 8B-F). Light stimulus in these neurons caused a
decrease in spontaneous action potential firing rate (Fig. 12D). In 10 out of 15 cells
expressing Rh-CT5-HT1A, the interspike interval was increased in response to a 3 s light stimulus on average from 202 ± 21 ms (n = 10) to 313 ± 58 ms and returned to 207 ± 25 ms, 13-16 seconds after cessation of light application (Fig. 12E and F). The change in firing rate could be attributed to enhancement of K+ conductance (most likely mediated
by GIRK channels) revealed by increased inward rectification with light (Figure 8B and
C). Eliciting light responses in brainstem slices required sufficient 9-cis retinal loading
with FAF-BSA supplemented extracellular solution to facilitate retinal delivery as
119
previously suggested (Li et al., 1999). These results indicate that Rh-CT5-HT1A can functionally replace 5-HT1A in the DRN neurons of 5-HT1A (-/-) mice.
120
Figures
Figure 1 - Vertebrate rhodopsin does not activate GIRK channels in the absence of light stimulus.
HEK293 (tsA201) cells were co-transfected with either GIRK1/4 subunits alone (A) or
GIRK1/4 and Rh (B) and currents were measured at a holding potential of -60 mV. 0.6 mM GTPγS present in the intracellular recording solution caused constitutive G protein activation and subsequent GIRK current enhancement. A low K+ (2 mM) solution was applied for 10 s (white bars) at 5 s and 5 min after establishment of the whole cell mode.
(C) Absolute inward currents through GIRK channels were calculated as the difference between current in normal (high K+) extracellular recording solution and low K+ (2 mM).
The current induced by GTPγS was calculated as the difference between absolute GIRK current at 5 min and 5 s. (D) Maximal GIRK current was determined by calculating induced GIRK current at 5 min after establishment of whole cell mode. Quantification of
GIRK current induced by GTPγS shows no significant difference between HEK293 cell transfected with Rh and GIRK1/4 or GIRK1/4 subunits alone. The number in parentheses indicate the number of experiments and statistical significance is noted as indicated (mean ± S.E.M).
121
122
Figure 2 - Amino acid sequence alignments of mCherry tagged GPCRs Rh, 5-HT1A
and Rh-CT5-HT1A
Protein alignment of wild-type GPCRs [rat rhodopsin (Rh) and human 5-HT1A] and the chimera Rh-CT5-HT1A using clustalW (Thompson et al., 1994). Grey highlighting
signifies highly conserved residues; yellow, the consensus E/DRY sequence; and green,
intracellular domains as previously predicted (Airan et al., 2009; Kim et al., 2005).
Transmembrane domains (TM), intracellular loops (I) and C-terminal domain (CT) are
marked. Primary sequence corresponding to mCherry is denoted by red font.
123
124
Figure 3 - Protein sequence alignments of Rh, 5-HT1A-mCherry and OptoXRs
Protein alignment of wild-type GPCRs [rat rhodopsin (Rh) and human 5-HT1A] and the chimeric OptoXRs (Opto-5-HT1A, Opto-β2AR, Opto-α1AR) Rh-CT5-HT1A using clustalW
(Thompson et al., 1994). Grey highlighting signifies highly conserved residues; yellow, the consensus E/DRY sequence; and green, intracellular domain sequences exchanged for chimera construction as noted previously (Airan et al., 2009; Kim et al., 2005).
Transmembrane domains (TM), intracellular loops (I) and C-terminal domain (CT) are
marked.
125
126
Figure 4 - Opto-5-HT1A is functional in HEK293 cells, but exhibits atypical GPCR
activation and fails to inactivate
(A) Schematic representations of the GPCR, Opto-5-HT1A. The chimera, contains the three intracellular domains and the C-terminal domain of human 5-HT1A receptor and is
tagged C-terminally with the fluorescent tag, mCherry. (B) "Short" time course
recording of a HEK293 cell co-transfected with GIRK 1/4 subunits and Opto-5-HT1A.
shows slow GIRK current induction upon a 10 sec pulse of light (485 nm). A low K+ (2 mM) solution was applied for 10 s (white bars) at 5 s and then after 2.5 min signified by the transient decrease in inward current. Absolute inward through GIRK channels, calculated as the difference between current in normal (high K+) extracellular recording
solution and low K+ (2 mM) confirms GIRK current activation. (C) "Long," 20 min
whole cell patch clamp recording of a HEK293 cell co-transfected with GIRK 1/4
subunits and Opto-5-HT1A. Repeated 10 sec pulse of light (485 nm) interspersed by 2.5
min dark periods induced maximal GIRK current. 10 s application of low K+ (2 mM)
solution at 5 s and then every 2.5 mins demonstrates gradual increase in GIRK mediated
K+ current.
127
128
Figure 5 - Functional expression and characterization of Rh-CT5-HT1A in HEK293
cells.
(A) Schematic representations of GPCRs C-terminally tagged with mCherry used for
exogenous expression. The chimera Rh-CT5-HT1A, contains the C-terminal domain of
human 5-HT1A receptor after the fluorescent tag. (B) Colocalization of GPCRs with
GIRK channels exogenously expressed in HEK293 cells. Cells were co-transfected with mCherry tagged receptors, GIRK1 subunit, and GIRK4 subunit tagged N-terminally with
EGFP. GPCRs (left, red) and GIRK1/4 channels (center, green) target efficiently to the cell membrane. (Right) Overlay of left and right panels shows colocalization of transfected GPCRs and GIRK channels indicated by yellow color (Scale bar = 10 µm).
Time course of GPCR-induced GIRK current demonstrate that activation by Rh-mCherry and Rh-CT5-HT1A is comparable to the GIRK current mediated by 5-HT1A and 5-HT1A- mCherry. Light sensitive GPCRs (C-D) were activated with a 10 s light pulse (485 nm) and 1 µM 8OH-DPAT was applied to cells transfected with 5-HT1A receptors (E-F).
Currents were measured at a holding potential of -60 mV . (G) Average light and agonist
induced GIRK current amplitude. (H) Comparison of the time constants of the GPCR-
induced GIRK current before and after GPCR activation. Number in parentheses indicate
the number of experiments and statistical significance is noted as indicated (mean ±
S.E.M; * significantly different from Rh, p < 0.05; ** significantly different from Rh, p <
0.001; ANOVA, Tukey's post-hoc test).
129
130
Figure 6 - Decremental decrease in GIRK channel activation by Rh in HEK cells with repeated light stimulation
(A) Time course of GPCR-induced GIRK current increase in HEK293 cells co- expressing Rh-mCherry and GIRK channel subunits. GIRK current induction by successive light stimulations decreases relative to initial (light-naive) response. 10 s light pulses of 485 nm (green bars) were applied three times with 1 minute dark periods interspersed to allow for inactivation. GIRK currents were measured at a holding potential of -60 mV. (B) Normalized light induced GIRK current increase with successive light applications. tsA201 cells were co-transfected with GIRK channel and untagged Rh, Rh-mCherry or Rh-CT5-HT1A. The GIRK current induced by second and third light pulses were normalized to current induced by initial stimulation. (mean ±
S.E.M; responses for 2nd and 3rd light applications for all constructs tested were significantly different from initial response (1.0), p < 0.001)
131
132
Figure 7 - The C-terminal domain of 5-HT1A is sufficient to induce targeting of vertebrate rhodopsin somatodendritically in neurons.
Confocal immunofluorescence images were taken of cultured rat hippocampal neurons
(8 DIV) infected with Sindbis virus driving expression of Rh-mCherry (A-H), Rh-CT5-
HT1A (I-L) and 5-HT1A-mCherry (M-P). Images are representative z-stack images projected to 2 dimensions. All Sindbis virus vectors also induced expression of EGFP
(B, F, J, N) under the control of a second subgenomic promoter. Neurons were stained with anti-dsRed antibody (A, E, I, M) to delineate the distribution of mCherry tagged receptors and were colabeled with the dendritic marker, MAP-2 (C) or axonal marker
Tau-1 (G, K, O). (A-D) Rh-mCherry targeted to both axons and dendrites in cultured neurons. Neurons infected with Rh-mCherry virus showed processes with dsRed, GFP and MAP-2 staining. Rh-mCherry and GFP were also found in processes that lacked
MAP-2 expression, labeled by white arrows. Overlay of dsRed, GFP and MAP-2
staining showed colocalization pattern. (E-H) Rh-mCherry targeting to axons was revealed by the presence of processes colabeled with anti-dsRed and anti-Tau-1
antibodies, indicated by white arrows. (I-L) Rh-CT5-HT1A expressed in hippocampal
neurons was present in dendrites, but was not targeted to GFP positive and Tau-1 positive
processes. White arrows mark examples of these axons. (M-P) 5-HT1A-mCherry was
also absent in GFP positive and Tau-1 positive processes. White arrows mark one of
these axons. Scale bars represents 20 µm.
133
134
Figure 8 - The C-terminal domain of 5-HT1A receptor promotes distal targeting within dendrites of hippocampal neurons.
Confocal images were taken of cultured rat hippocampal neurons (8 DIV) infected with
Sindbis virus driving expression of Rh-mCherry (A), Rh-CT5-HT1A (B) and 5-HT1A-
mCherry (C). These neurons were immunolabeled with anti-dsRed (red) and anti-MAP-2
(not shown) antibodies to enhance and delineate mCherry tagged receptors and dendrites,
respectively. Representative z-stack images taken at lower (20x) magnification were
projected to 2 dimensions. These images reveal that dsRed fluorescence is observed
much more distally from the soma for neurons expressing Rh-CT5-HT1A and 5-HT1A-
mCherry when compared to neurons expressing Rh-mCherry. Scale bar represents 20
µm. (E) Normalized fluorescence of the longest dendrite of a given neuron was
quantified as a function of distance from the soma. Fluorescence plots versus absolute
distance for Rh-mCherry (blue), Rh-CT5-HT1A (red) and 5-HT1A-mCherry (green) are
shown (mean ± S.E.M; n = 16) (F) Rh-CT5-HT1A and 5-HT1A-mCherry target further along the extent of dendrites. Normalized fluorescence of the longest dendrite was plotted against normalized dendritic length. The length of each dendrite analyzed was normalized from 0 to 1.0 by piecewise linear interpolation. Interpolated data were pooled and mean ± S.E.M plotted against normalized distance (n = 16). (D) Lengths of dendrites analyzed were not significantly different for each cell condition.
135
136
Figure 9 - Rh-CT5-HT1A induces membrane hyperpolarization in rat hippocampal
neurons with light stimulus.
Hyperpolarization induced by Rh-mCherry (A) and Rh-CT5-HT1A (B) in cultured hippocampal neurons (14 DIV) during a 1 s pulse of 485 nm light. Light activated receptor-induced voltage change was comparable to agonist induced, Gi/o-linked GPCR activation. (C) Average hyperpolarization induced by 1 s light stimulus (Rh-mCherry
and Rh-CT5-HT1A), application of 50 μM baclofen, or application of 1 μM of 8OH-DPAT in cultured rat hippocampal neurons (10-14 DIV). For recordings with light stimulus, neurons were infected with Sindbis virus driving expression of corresponding GPCR.
Agonist induced responses were determined in uninfected neurons. (D) Time course of
GPCR (Rh-mCherry, Rh-CT5-HT1A, GABAB, or 5-HT1A)-induced hyperpolarization and recovery from hyperpolarization after switching off the light or washing out agonist.
Average change in membrane potential induced and activation and inactivation time constants were not significantly different from each other (p > 0.05, ANOVA).
137
138
Figure 10 - Rh-CT5-HT1A replaces endogenous 5-HT1A receptors in hippocampal
neurons.
Rh-CT5-HT1A but not Rh-mCherry decreases endogenous 5HT1A-induced
hyperpolarization without affecting GABAB responses. (A) Extent of membrane
hyperpolarization induced by 5-HT1A activation is decreased in neurons expressing Rh-
CT5-HT1A, but not Rh-mCherry. Cultured rat hippocampal neurons (21-22 DIV) were
infected with Sindbis virus driving the expression of Rh-mCherry or Rh-CT5-HT1A.
Voltage changes induced by a 10 s application of the 5HT1A agonist, 8OH-DPAT (1µM),
in the presence of Rh-CT5-HT1A (top) and Rh-mCherry (bottom). (B) Relative changes in
membrane voltage for 5HT1A (8OH-DPAT) and GABAB (baclofen) activation in
uninfected (WT) neurons compared to those expressing Rh-CT5-HT1A or Rh-mCherry (** significance with respect to WT and Rh expressing neurons, p < 0.01; ANOVA, Tukey's post-hoc test).
139
140
Figure 11 - Light activation of Rh-CT5-HT1A functionally rescues 5-HT1A loss-of-
function phenotype in cultured hippocampal neurons.
(A) Confocal images confirm the lack of 5-HT1A receptors in 5-HT1A KO mice (center- left column). Neurons from wild type (left column) and 5-HT1A KO mice (9 DIV) were
immunolabeled with anti-5-HT1A (upper, red) and anti-MAP-2 (middle, green) antibodies. (Lower, left panels) Overlay shows significant colocalization of 5-HT1A and
MAP-2 in wild type neurons (yellow). Virally induced 5-HT1A-mCherry and Rh-CT5-
HT1A were expressed in the dendrites of cultured hippocampal neurons of 5-HT1A null
mice (center-right and right columns). Virally transfected KO neurons were stained with
anti-dsRed and MAP-2 antibodies to visualize distribution of mCherry tagged GPCRs
and dendrites, respectively. (Lower, right panels) Overlay of dsRed and MAP-2 staining
reveals a high degree of colocalization of virally transfected 5-HT1A-mCherry and Rh-
CT5-HT1A with MAP-2. (B and C) Voltage change induced in cultured hippocampal
neurons from wild type mice by baclofen (B) and 8OH-DPAT (C). (D and E)
Membrane hyperpolarization is induced in cultured hippocampal neurons from 5-HT1A
null mice by baclofen (D), but not 8OH-DPAT (E). (F and G) Agonist (8OH-DPAT)
activation of 5-HT1A-mCherry (F) and light stimulation of Rh-CT5-HT1A (G) expressed in
5-HT1A KO neurons induce hyperpolarization in a similar pattern to the response of wild type neurons to 8OH-DPAT application. (H) Comparison of the hyperpolarization induced by baclofen, 8OH-DPAT, or light stimulus between KO and wild type neurons in the presence or absence of 5-HT1A-mCherry or Rh-CT5-HT1A. (I) Time course of GPCR
(GABAB, 5-HT1A, 5-HT1A-mCherry or Rh-CT5-HT1A)-induced hyperpolarization and
141
recovery from hyperpolarization after switching off the light or washing out agonist in wild type (WT) and KO mouse neurons. (n.s., p > 0.05, ANOVA).
142
143
Figure 12 - Rh-CT5-HT1A compensates for loss of 5-HT1A mediated signaling of
neurons in the dorsal raphe nuclei of 5-HT1A KO mice.
(A) Functional expression of Rh-CT5-HT1A in 5-HT neurons of the dorsal raphe.
Intracranial injections into the dorsal raphe were performed on ePet::YFP transgenic
mice. A Lentiviral vector drove the expression of Rh-CT5-HT1A under the control of a
CMV promoter (left). YFP expressed under the control of a serotonergic specific promoter, ePet-1, labeled 5-HT neurons (right). Punctate distribution of Rh-CT5-HT1A was
observed in neurons 9 days after injection. DR slices were stained with anti-GFP
antibody to amplify the YFP signal. Rh-CT5-HT1A expression functionally rescued loss-
of-function phenotype in 5-HT neurons of KO mice. (B) Current traces elicited in DRN
neurons from 5-HT1A KO mice expressing Rh-CT5-HT1A by a voltage ramp from -120 to -
45 mV. Light application (490 nm) increased membrane currents. (C) Quantification of
the light-induced current measured at -120 mV. (D) (Top) Spontaneous action potential
firing of DRN neurons from 5-HT1A null mice expressing Rh-CT5-HT1A was reduced by a
3 s light pulse (490 nm). (Bottom) During the light pulse the interspike interval was
increased during the light pulse and decreased to resting levels once the light was
switched off. (E) Plot of the interspike interval for a single experiments of DRN neurons
from 5-HT1A (-/-) mice expressing Rh-CT5-HT1A before, during and after a 3 s, 490 nm
light pulse. (F) Percent change of the interspike interval for 5-HT1A (-/-) DRN neurons expressing Rh-CT5-HT1A before, during and after a 3 s, 490 nm light pulse.
144
145
CHAPTER 6
DISCUSSION
146
Research conclusions
In this study, we have described the construction and characterization of a chimeric light-activated receptor that targets and functions in 5-HT1A receptor signaling
domains. With the addition of mCherry and the CT domain of 5-HT1A, vertebrate
rhodopsin retains its ability to activate Gi/o-coupled signaling, causes subsequent GIRK
channel activation, and induces membrane hyperpolarization in neurons. Importantly, in
comparison to agonist induced wild type 5-HT1A activity, Rh-CT5-HT1A induces GIRK
current and membrane hyperpolarization to a statistically indistinguishable degree. The
kinetics of activation and inactivation of these responses are also similar indicating that
there is functional homology between the chimera and 5-HT1A. Although in HEK cells,
GIRK current induction occurs faster for 5-HT1A in comparison to light sensitive
receptors, this effect is not observed in neurons presumably due to a different milieu of G
protein modulators present in neurons. The necessity to include the CT tag in the
chimera to mimic natural 5-HT1A is evident when examining intracellular targeting.
When expressed in neurons, Rh-CT5-HT1A traffics to somatodendritic compartments and to
distal dendritic segments, where endogenous and exogenously expressed 5-HT1A
receptors are found. This is in contrast to Rh lacking the CT tag, which targets axonally
and somatodendritically and exhibits stunted trafficking along dendrites. Finally, Rh-
CT5-HT1A rescues the cellular loss of function phenotype in both cultured hippocampal
neurons and neurons of the dorsal raphe in hindbrain slices indicating that Rh-CT5-HT1A
can act in place of endogenous 5-HT1A receptor. Taken together, these findings suggests
that light activation of Rh-CT5-HT1A serves as a suitable proxy for agonist-induced 5-HT1A
receptor activation.
147
In developing Rh-CT5-HT1A, we have addressed some practical considerations for using light activated probes in vitro. Since photoactive vitamin A derivatives can be present in culture media and Rh can be weakly activated and rendered refractory even without exogenous retinal supplementation, fluorescent tags excited by similar wavelength to that of the light activated probe should be avoided. In addition, to prevent
inadvertent receptor activation, exposure to ambient light should be minimized. We have
also shown in this study that exogenously expressed Rh is not active in the absence of
light when kept in dark conditions. This is of critical importance because any
background constitutive activity would weaken the validity of the use of heterologously
expressed Rh as a optical, molecular switch. If this were the case, the mere expression of
the light sensitive probe would affect the steady state properties of neurons. Although
light activated receptors have an intrinsic experimental control state (light off), modeling
of in vivo 5-HT1A activity during normal and diseased behavior would be skewed.
Critical trafficking domains as molecular tags to direct intracellular targeting
We have also demonstrated that the intracellular trafficking of heterologously
expressed proteins can be directed to specific subcellular domains by the addition of
critical targeting domains of other proteins. For Rh-CT5-HT1A, an important consequence
of CT directed targeting away from axons is the elimination of its potential influence on
presynaptic modulators of vesicle release. In addition to its effects on GIRK channels
and membrane voltage, Rh inhibits presynaptic P/Q-type Ca2+ channel currents which
then causes increases in paired pulse facilitation ratio (Li et al., 2005a). This
complication should be eliminated so that Rh-CT5-HT1A activation in neurons is limited to
148
somatodendritic GIRK current modulation and not the quantal content of vesicle release.
The differences between the 5-HT1A and 5-HT1B are illustrative of this point. Both 5-
HT1A and 5-HT1B are similar with respect to amino acid homology and downstream G
protein signaling by coupling negatively to adenylate cyclase. However, 5-HT1A
modulates neuronal firing, whereas 5-HT1B participates in the local control of
neurotransmitter release from the terminal (Sari, 2004). These differences in function can
be attributed to the respective localization patterns of the receptors. More broadly,
directed subcellular targeting of exogenously expressed proteins could be applied to other
receptor systems. However, one needs to be careful not to exclude the possibility that
other intracellular domains within Rh cooperate with CT to direct somatodendritic
targeting. Thus, the successful application of this type of strategy may depend on the presence of specific binding motifs on the donor receptor in addition to the targeting tag.
Our efforts to further replace the intracellular domains of Rh using the corresponding 5-
HT1A receptor domains produced a light activated chimeric GPCR with unusual
activation and inactivation kinetics (Chapter 5, Fig. 4). Since the domain predictions
were derived from the published chimeric receptors between rhodopsin and adrenergic
receptors (Airan et al., 2009; Kim et al., 2005) the results suggest that the intracellular protein domains between 5-HT and adrenergic GPCRs are divergent.
Endogenous receptor replacement by exogenous receptor expression
Tagging with critical targeting domains to drive differential intracellular
localization also has the potential to induce competitive substitutions of the endogenously
expressed proteins. The strategy of adding targeting domain of exogenous receptors
149
depends on their interaction with chaperone proteins normally binding and trafficking
with endogenous receptors. Therefore, exogenously expressed proteins expressed at high
enough levels would create an analogous situation to overexpression of the targeting
domain alone. This could result in a dominant negative effect by direct competition with
endogenous receptor. We reason that this would occur since there are a finite number of trafficking proteins and a finite number of positions GPCRs can occupy at a given submembrane locale. Since our goal is to utilize Rh-CT5-HT1A as a functional substitute 5-
HT1A, the pseudo-knockdown of endogenous 5-HT1A signaling would be desirable. For correlative and causal studies linking 5-HT1A-like signaling of Rh-CT5-HT1A to behavior,
the compensatory effects from endogenous 5-HT1A signaling would be minimized.
Remaining questions and future directions
Because Rh-CT5-HT1A can serve as a functional substitute for endogenous 5-HT1A,
the chimera could prove to be a powerful tool for defining in vivo 5-HT1A function in
future studies. For example, expression of Rh-CT5-HT1A in 5-HT1A null animals will
potentially reveal the levels and temporal patterns of 5-HT1A signaling sufficient to
restore normal behavior. Comparison of Rh-CT5-HT1A activation in constitutional null
(Heisler et al., 1998) and conditional knockdown animals (Richardson-Jones et al., 2010)
could be used to differentiate between the consequences of 5-HT1A function during
developmental and adulthood in serotonergic neurons versus the peripheral targets. Are
the physiological effects exerted by SSRI therapy a cell-autonomous phenomenon restricted specifically to modulation of 5-HT activity and subsequent regulation of serotonin tone? Functional expression of the chimera in wild type and conditional null
150
animals will elucidate consequences of excess 5-HT1A activity and could clarify the
relationship between 5-HT1A signaling, receptor auto-regulation and SSRI treatment
efficacy.
Indeed, the light activated chimera would be useful for defining 5-HT1A
autoreceptor function, but could be expressed outside of the 5-HT system to examine 5-
HT1A signaling in non-5HT neurons and peripheral targets. As heteroreceptors, 5-HT1A
receptors are widely expressed in the CNS and periphery. 5-HT1A is expressed in
lymphatic tissue, gut, muscle and kidney (Albert et al., 1990; Kobilka et al., 1987).
Within the CNS brain, 5-HT1A is highly enriched in the hippocampus, entorhinal cortex,
and raphe nuclei, with lower levels found in neocortex and thalamus (Chalmers and
Watson, 1991).
One of the physiological roles of serotonin signaling is the modulation of breathing behaviors. The neuronal circuits that control breathing (brainstem central pattern generators and sensory afferent networks) receive significant input from serotonergic neurons. Interestingly, serotonergic dysfunction has been implicated in the pathogenesis of sudden infant death syndrome. Widespread decrease in 5-HT1A agonist
binding and aberrant 5-HT1A expression throughout the hindbrain is observed in SIDS
victims (Ozawa and Okado, 2002; Paterson et al., 2006). Given the likelihood that 5-
HT1A function is an important regulator of breathing behavior, we have analyzed the 5-
HT1A knockout mice for inherent differences. These preliminary experiments were
conducted in conjunction with the laboratory of Dr. Ted Dick (Case Western Reserve
University, Department of Medicine). Wild type and 5-HT1A (-/-) knock out animals
were anaesthetized and their breathing responses were measured by plethysmography.
151
This is a tractable behavioral study since the mice are anaesthetized during breathing and
no animal training/learning protocols are required. During hypoxic (low oxygen)
challenge, rodents respond in a stereotyped manner with a marked increase in respiratory
rate and tidal volume. The respiratory drive decreases upon re-oxygenation, but there is a
transient undershoot called the post-hypoxic frequency decline (PHFD) where the respiratory rate is lower than even basal levels. Preliminary data show that the PHFD is significantly attenuated or missing in knockout animals. Since there is a clear behavioral phenotype, it would be interesting to see if exogenously expressed 5-HT1A and chimeric
receptor could functionally rescue this phenotype. The brainstems of 5-HT1A knockout
mice would be injected with a Lentivirus vectors expressing fluorescently tagged 5-HT1A
or Rh-CT5-HT1A, and then 7-9 days later, breathing behavior again could be analyzed by
plethysmography. Virally delivered receptors will then be stimulated by either cannular
application of 8OH-DPAT (5-HT1A agonist) or an implanted light-guide. Post-hoc
imaging would confirm correct regional expression of receptors. The ability of chimeric
receptors to rescue the knockout phenotype will be compared with that of degree of
rescue with native 5-HT1A receptor. Successful rescue would be of great significance
because it would indicate that Rh-CT5-HT1A could functionally substitute for 5-HT1A
regulation of behavior in vivo. Furthermore, post-hoc immunofluorescence in animals
with restored breathing (compared to those with no rescue) could be used to start
delineating which neurons are involved in the 5-HT1A-mediated control of breathing.
As a more general question, outside of central serotonergic neurons, it would be very interesting to start examining the relationship between serotonin signaling and somatic manifestations of anxiety and depression (such as changes in sleep and
152
gastrointestinal function). The main question to address would be to determine if somatic
symptoms of psychiatric disease are purely secondary to physiologic stress responses
(mediated by cortisol, growth hormone and norepinephrine) or if their etiology directly
involves changes in serotonin signaling. These questions would be answered by
expressing Rh-CT5-HT1A in other brain regions and peripheral tissues that normally
express 5-HT1A, and then determining how stimulation of these receptors correlates to
physiologic response. Then parallel experiments would be performed in animal models
of anxiety and depression (i.e. 5-HT1A null mouse) to assess if Rh-CT5-HT1A can
functionally rescue behavioral phenotypes associated with disease. For example, 5-HT1A
receptors are enriched in orexin/hypocretin neurons of the lateral hypothalamus (Muraki
et al., 2004). Orexin neurons are critical for the regulation of sleep/wake cycle receive and serotonergic input. 5-HT then hyperpolarizes orexin neurons by activating the 5-
HT1A receptor (Muraki et al., 2004). An optogenetic approach has been taken to study the function of hypocretin neurons, where channelrhopdopsin-2 was expressed specifically in orexin neurons using a Hrct promoter (Adamantidis et al., 2007). Light
application caused depolarization and action potential firing in orexin neurons. The in
vivo behavioral consequence was an increased probability of transition to wakefulness
(Adamantidis et al., 2007). A similar optogenetic experiment could be performed to
specifically implicate 5-HT1A signaling in the regulation of orexin neurons and the
regulation of sleep-wakefulness. Activation of Rh-CT5-HT1A would presumably silence
hypocretin neuron firing and most likely inhibit transition to wakefulness. 5-HT1A
knockout mice could then be analyzed for defects in sleep-wake cycle and then functional rescue of these phenotypes by Rh-CT5-HT1A chimera would be attempted. The ideal
153
premise of these studies is demonstration of the sufficiency and/or necessity of 5-HT1A signaling with respect to specific behaviors and physiological responses. Generally speaking, this may lead to new therapies for the symptomatic treatment of mental illness that complements SSRI mediated modulation of central serotonin release. This could provide positive cognitive feedback that may be very beneficial for patients' recovery.
Unfortunately, our attempts to clone and characterize a chimeric receptor containing the intracellular domains of the 5-HT1A receptor lead to synthesizing a GPCR that activated Gi/o signaling in response to light, but enhanced GIRK current very slowly and was constitutively active. To generate a GPCR with kinetic properties more closely resembling wild type GPCRs, we would need to better define transmembrane to intracellular domain boundaries for 5-HT1A. This would involve a more systematic approach where only single intracellular domains would be exchanged at each iteration and different combinations of amino acids retained/replaced would be assessed. This is similar to the approach taken by the Khorana lab (Kim et al., 2005). For whatever reason, there may be complication in converting a Gi/o linked receptor to another Gi/o couple receptor. Successful exchange of intracellular domains has been documented for
Rh (Gi/o) conversion to Rh/β2-adrenergic (Gs) and Rh/α1-adrenergic (Gq) receptors.
Therefore, another approach that could be taken in which invertebrate Rh could be used as the GPCR backbone, and the intracellular domains of squid Rh could be exchanged for those of 5-HT1A. Crystal structure of squid rhodopsin has recently been solved and reveals a peculiar features that distinguishes it from vertebrate rhodopsin and may be crucial for Gq coupling (Murakami and Kouyama, 2008; Palczewski et al., 2000; Teller et al., 2001). This may facilitate construction of a full intracellular chimera because wild
154
type squid Rh should not activate Gi/o coupled downstream targets, but a successfully
converted chimera containing intracellular 5-HT1A domain would presumable activate
GIRK. In this way, each iteration of the chimera could be assessed very clearly for ability to Gi/o signaling.
155
REFERENCES
Aberle, H., A.P. Haghighi, R.D. Fetter, B.D. McCabe, T.R. Magalhaes, and C.S.
Goodman. 2002. wishful thinking encodes a BMP type II receptor that regulates
synaptic growth in Drosophila. Neuron. 33:545-558.
Adamantidis, A.R., F. Zhang, A.M. Aravanis, K. Deisseroth, and L. de Lecea. 2007.
Neural substrates of awakening probed with optogenetic control of hypocretin
neurons. Nature. 450:420-424.
Adams, M.D., S.E. Celniker, R.A. Holt, C.A. Evans, J.D. Gocayne, P.G. Amanatides,
S.E. Scherer, P.W. Li, R.A. Hoskins, R.F. Galle, R.A. George, S.E. Lewis, S.
Richards, M. Ashburner, S.N. Henderson, G.G. Sutton, J.R. Wortman, M.D.
Yandell, Q. Zhang, L.X. Chen, R.C. Brandon, Y.H. Rogers, R.G. Blazej, M.
Champe, B.D. Pfeiffer, K.H. Wan, C. Doyle, E.G. Baxter, G. Helt, C.R. Nelson,
G.L. Gabor, J.F. Abril, A. Agbayani, H.J. An, C. Andrews-Pfannkoch, D.
Baldwin, R.M. Ballew, A. Basu, J. Baxendale, L. Bayraktaroglu, E.M. Beasley,
K.Y. Beeson, P.V. Benos, B.P. Berman, D. Bhandari, S. Bolshakov, D. Borkova,
M.R. Botchan, J. Bouck, P. Brokstein, P. Brottier, K.C. Burtis, D.A. Busam, H.
Butler, E. Cadieu, A. Center, I. Chandra, J.M. Cherry, S. Cawley, C. Dahlke, L.B.
Davenport, P. Davies, B. de Pablos, A. Delcher, Z. Deng, A.D. Mays, I. Dew,
S.M. Dietz, K. Dodson, L.E. Doup, M. Downes, S. Dugan-Rocha, B.C. Dunkov,
P. Dunn, K.J. Durbin, C.C. Evangelista, C. Ferraz, S. Ferriera, W. Fleischmann,
C. Fosler, A.E. Gabrielian, N.S. Garg, W.M. Gelbart, K. Glasser, A. Glodek, F.
Gong, J.H. Gorrell, Z. Gu, P. Guan, M. Harris, N.L. Harris, D. Harvey, T.J.
Heiman, J.R. Hernandez, J. Houck, D. Hostin, K.A. Houston, T.J. Howland, M.H.
156
Wei, C. Ibegwam, et al. 2000. The genome sequence of Drosophila melanogaster.
Science. 287:2185-2195.
Adell, A., P. Celada, M.T. Abellan, and F. Artigas. 2002. Origin and functional role of
the extracellular serotonin in the midbrain raphe nuclei. Brain Res Brain Res Rev.
39:154-180.
Adler, A.J., and R.B. Edwards. 2000. Human interphotoreceptor matrix contains serum
albumin and retinol-binding protein. Exp Eye Res. 70:227-234.
Airan, R.D., K.R. Thompson, L.E. Fenno, H. Bernstein, and K. Deisseroth. 2009.
Temporally precise in vivo control of intracellular signalling. Nature. 458:1025-
1029.
Albert, P.R., Q.Y. Zhou, H.H. Van Tol, J.R. Bunzow, and O. Civelli. 1990. Cloning,
functional expression, and mRNA tissue distribution of the rat 5-
hydroxytryptamine1A receptor gene. J Biol Chem. 265:5825-5832.
Alenina, N., D. Kikic, M. Todiras, V. Mosienko, F. Qadri, R. Plehm, P. Boye, L.
Vilianovitch, R. Sohr, K. Tenner, H. Hortnagl, and M. Bader. 2009. Growth
retardation and altered autonomic control in mice lacking brain serotonin. Proc
Natl Acad Sci U S A. 106:10332-10337.
Andersson, F., J. Jakobsson, P. Low, O. Shupliakov, and L. Brodin. 2008. Perturbation of
syndapin/PACSIN impairs synaptic vesicle recycling evoked by intense
stimulation. J Neurosci. 28:3925-3933.
Anggono, V., and P.J. Robinson. 2007. Syndapin I and endophilin I bind overlapping
proline-rich regions of dynamin I: role in synaptic vesicle endocytosis. J
Neurochem.
157
Anggono, V., K.J. Smillie, M.E. Graham, V.A. Valova, M.A. Cousin, and P.J. Robinson.
2006. Syndapin I is the phosphorylation-regulated dynamin I partner in synaptic
vesicle endocytosis. Nat Neurosci. 9:752-760.
Artigas, F., V. Perez, and E. Alvarez. 1994. Pindolol induces a rapid improvement of
depressed patients treated with serotonin reuptake inhibitors. Arch Gen
Psychiatry. 51:248-251.
Ase, A.R., T.A. Reader, R. Hen, M. Riad, and L. Descarries. 2001. Regional changes in
density of serotonin transporter in the brain of 5-HT1A and 5-HT1B knockout
mice, and of serotonin innervation in the 5-HT1B knockout. J Neurochem.
78:619-630.
Ataman, B., J. Ashley, D. Gorczyca, M. Gorczyca, D. Mathew, C. Wichmann, S.J.
Sigrist, and V. Budnik. 2006. Nuclear trafficking of Drosophila Frizzled-2 during
synapse development requires the PDZ protein dGRIP. Proc Natl Acad Sci U S A.
103:7841-7846.
Audero, E., E. Coppi, B. Mlinar, T. Rossetti, A. Caprioli, M.A. Banchaabouchi, R.
Corradetti, and C. Gross. 2008. Sporadic autonomic dysregulation and death
associated with excessive serotonin autoinhibition. Science. 321:130-133.
Ball, R.W., M. Warren-Paquin, K. Tsurudome, E.H. Liao, F. Elazzouzi, C. Cavanagh,
B.S. An, T.T. Wang, J.H. White, and A.P. Haghighi. 2010. Retrograde BMP
signaling controls synaptic growth at the NMJ by regulating trio expression in
motor neurons. Neuron. 66:536-549.
158
Bantick, R.A., J.F. Deakin, and P.M. Grasby. 2001. The 5-HT1A receptor in
schizophrenia: a promising target for novel atypical neuroleptics? J
Psychopharmacol. 15:37-46.
Banzai, Y., H. Miki, H. Yamaguchi, and T. Takenawa. 2000. Essential role of neural
Wiskott-Aldrich syndrome protein in neurite extension in PC12 cells and rat
hippocampal primary culture cells. J Biol Chem. 275:11987-11992.
Barnes, N.M., and T. Sharp. 1999. A review of central 5-HT receptors and their function.
Neuropharmacology. 38:1083-1152.
Barnes, P.J., K.F. Chung, and C.P. Page. 1998. Inflammatory mediators of asthma: an
update. Pharmacol Rev. 50:515-596.
Bateman, J.R., A.M. Lee, and C.T. Wu. 2006. Site-specific transformation of Drosophila
via phiC31 integrase-mediated cassette exchange. Genetics. 173:769-777.
Bekkers, J.M., and C.F. Stevens. 1991. Excitatory and inhibitory autaptic currents in
isolated hippocampal neurons maintained in cell culture. Proc Natl Acad Sci U S
A. 88:7834-7838.
Bellen, H.J., R.W. Levis, G. Liao, Y. He, J.W. Carlson, G. Tsang, M. Evans-Holm, P.R.
Hiesinger, K.L. Schulze, G.M. Rubin, R.A. Hoskins, and A.C. Spradling. 2004.
The BDGP gene disruption project: single transposon insertions associated with
40% of Drosophila genes. Genetics. 167:761-781.
Ben-Yaacov, S., R. Le Borgne, I. Abramson, F. Schweisguth, and E.D. Schejter. 2001.
Wasp, the Drosophila Wiskott-Aldrich syndrome gene homologue, is required for
cell fate decisions mediated by Notch signaling. J Cell Biol. 152:1-13.
159
Bennett, V., and A.J. Baines. 2001. Spectrin and ankyrin-based pathways: metazoan
inventions for integrating cells into tissues. Physiol Rev. 81:1353-1392.
Berger, S., G. Schafer, D.A. Kesper, A. Holz, T. Eriksson, R.H. Palmer, L. Beck, C.
Klambt, R. Renkawitz-Pohl, and S.F. Onel. 2008. WASP and SCAR have distinct
roles in activating the Arp2/3 complex during myoblast fusion. J Cell Sci.
121:1303-1313.
Blier, P., C. de Montigny, and Y. Chaput. 1987. Modifications of the serotonin system by
antidepressant treatments: implications for the therapeutic response in major
depression. J Clin Psychopharmacol. 7:24S-35S.
Bogdan, S., O. Grewe, M. Strunk, A. Mertens, and C. Klambt. 2004. Sra-1 interacts with
Kette and Wasp and is required for neuronal and bristle development in
Drosophila. Development. 131:3981-3989.
Boyden, E.S., F. Zhang, E. Bamberg, G. Nagel, and K. Deisseroth. 2005. Millisecond-
timescale, genetically targeted optical control of neural activity. Nat Neurosci.
8:1263-1268.
Bramham, C.R. 2008. Local protein synthesis, actin dynamics, and LTP consolidation.
Curr Opin Neurobiol. 18:524-531.
Brand, A.H., and N. Perrimon. 1993. Targeted gene expression as a means of altering cell
fates and generating dominant phenotypes. Development. 118:401-415.
Broadie, K. 2004. Synapse scaffolding: intersection of endocytosis and growth. Curr
Biol. 14:R853-855.
Broadie, K., and M. Bate. 1993a. Activity-dependent development of the neuromuscular
synapse during Drosophila embryogenesis. Neuron. 11:607-619.
160
Broadie, K., and M. Bate. 1993b. Innervation directs receptor synthesis and localization
in Drosophila embryo synaptogenesis. Nature. 361:350-353.
Broadie, K., and M. Bate. 1993c. Muscle development is independent of innervation
during Drosophila embryogenesis. Development. 119:533-543.
Broadie, K.S., and M. Bate. 1993d. Development of the embryonic neuromuscular
synapse of Drosophila melanogaster. J Neurosci. 13:144-166.
Brueggemann, L.I., and J.M. Sullivan. 2002. HEK293S cells have functional retinoid
processing machinery. J Gen Physiol. 119:593-612.
Budnik, V. 1996. Synapse maturation and structural plasticity at Drosophila
neuromuscular junctions. Curr Opin Neurobiol. 6:858-867.
Budnik, V., M. Gorczyca, and A. Prokop. 2006. Selected methods for the anatomical
study of Drosophila embryonic and larval neuromuscular junctions. Int Rev
Neurobiol. 75:323-365.
Budnik, V., Y. Zhong, and C.F. Wu. 1990. Morphological plasticity of motor axons in
Drosophila mutants with altered excitability. J Neurosci. 10:3754-3768.
Byk, T., M. Bar-Yaacov, Y.N. Doza, B. Minke, and Z. Selinger. 1993. Regulatory
arrestin cycle secures the fidelity and maintenance of the fly photoreceptor cell.
Proc Natl Acad Sci U S A. 90:1907-1911.
Carrel, D., M. Hamon, and M. Darmon. 2006. Role of the C-terminal di-leucine motif of
5-HT1A and 5-HT1B serotonin receptors in plasma membrane targeting. J Cell
Sci. 119:4276-4284.
161
Carrel, D., J. Masson, S. Al Awabdh, C.B. Capra, Z. Lenkei, M. Hamon, M.B. Emerit,
and M. Darmon. 2008. Targeting of the 5-HT1A serotonin receptor to neuronal
dendrites is mediated by Yif1B. J Neurosci. 28:8063-8073.
Cazzola, I., and M.G. Matera. 2000. 5-HT modifiers as a potential treatment of asthma.
Trends Pharmacol Sci. 21:13-16.
Chalmers, D.T., and S.J. Watson. 1991. Comparative anatomical distribution of 5-HT1A
receptor mRNA and 5-HT1A binding in rat brain--a combined in situ
hybridisation/in vitro receptor autoradiographic study. Brain Res. 561:51-60.
Chen, K., C. Merino, S.J. Sigrist, and D.E. Featherstone. 2005. The 4.1 protein coracle
mediates subunit-selective anchoring of Drosophila glutamate receptors to the
postsynaptic actin cytoskeleton. J Neurosci. 25:6667-6675.
Chintapalli, V.R., J. Wang, and J.A. Dow. 2007. Using FlyAtlas to identify better
Drosophila melanogaster models of human disease. Nat Genet. 39:715-720.
Cingolani, L.A., and Y. Goda. 2008. Actin in action: the interplay between the actin
cytoskeleton and synaptic efficacy. Nat Rev Neurosci. 9:344-356.
Clark, S.G., M.J. Stern, and H.R. Horvitz. 1992. C. elegans cell-signalling gene sem-5
encodes a protein with SH2 and SH3 domains. Nature. 356:340-344.
Co, C., D.T. Wong, S. Gierke, V. Chang, and J. Taunton. 2007. Mechanism of actin
network attachment to moving membranes: barbed end capture by N-WASP WH2
domains. Cell. 128:901-913.
Colino, A., and J.V. Halliwell. 1987. Differential modulation of three separate K-
conductances in hippocampal CA1 neurons by serotonin. Nature. 328:73-77.
162
Collins, C.A., and A. DiAntonio. 2004. Coordinating synaptic growth without being a
nervous wreck. Neuron. 41:489-491.
Collins, C.A., and A. DiAntonio. 2007. Synaptic development: insights from Drosophila.
Curr Opin Neurobiol. 17:35-42.
Collins, C.A., Y.P. Wairkar, S.L. Johnson, and A. DiAntonio. 2006. Highwire restrains
synaptic growth by attenuating a MAP kinase signal. Neuron. 51:57-69.
Cornwall, M.C., and G.L. Fain. 1994. Bleached pigment activates transduction in isolated
rods of the salamander retina. J Physiol. 480 ( Pt 2):261-279.
Cornwall, M.C., H.R. Matthews, R.K. Crouch, and G.L. Fain. 1995. Bleached pigment
activates transduction in salamander cones. J Gen Physiol. 106:543-557.
Coyle, I.P., Y.H. Koh, W.C. Lee, J. Slind, T. Fergestad, J.T. Littleton, and B. Ganetzky.
2004. Nervous wreck, an SH3 adaptor protein that interacts with Wsp, regulates
synaptic growth in Drosophila. Neuron. 41:521-534.
Dahmann, C. 2008. Drosophila : methods and protocols. Humana Press, Totowa, N.J. xi,
437 p. pp.
Daniels, R.W., C.A. Collins, K. Chen, M.V. Gelfand, D.E. Featherstone, and A.
DiAntonio. 2006. A single vesicular glutamate transporter is sufficient to fill a
synaptic vesicle. Neuron. 49:11-16.
Darmon, M., X. Langlois, L. Suffisseau, C.M. Fattaccini, and M. Hamon. 1998.
Differential membrane targeting and pharmacological characterization of
chimeras of rat serotonin 5-HT1A and 5-HT1B receptors expressed in epithelial
LLC-PK1 cells. J Neurochem. 71:2294-2303.
163
Davidson, R.J., K.M. Putnam, and C.L. Larson. 2000. Dysfunction in the neural circuitry
of emotion regulation--a possible prelude to violence. Science. 289:591-594.
Davis, G.W. 2006. Homeostatic control of neural activity: from phenomenology to
molecular design. Annu Rev Neurosci. 29:307-323.
Dawson, J.C., J.A. Legg, and L.M. Machesky. 2006. Bar domain proteins: a role in
tubulation, scission and actin assembly in clathrin-mediated endocytosis. Trends
Cell Biol. 16:493-498.
Dharmalingam, E., A. Haeckel, R. Pinyol, L. Schwintzer, D. Koch, M.M. Kessels, and B.
Qualmann. 2009. F-BAR proteins of the syndapin family shape the plasma
membrane and are crucial for neuromorphogenesis. J Neurosci. 29:13315-13327.
Dickman, D.K., Z. Lu, I.A. Meinertzhagen, and T.L. Schwarz. 2006. Altered synaptic
development and active zone spacing in endocytosis mutants. Curr Biol. 16:591-
598.
Dickopf, S., T. Mielke, and M.P. Heyn. 1998. Kinetics of the light-induced proton
translocation associated with the pH-dependent formation of the metarhodopsin
I/II equilibrium of bovine rhodopsin. Biochemistry. 37:16888-16897.
Ding, Y.Q., U. Marklund, W. Yuan, J. Yin, L. Wegman, J. Ericson, E. Deneris, R.L.
Johnson, and Z.F. Chen. 2003. Lmx1b is essential for the development of
serotonergic neurons. Nat Neurosci. 6:933-938.
Doumanov, J.A., M. Daubrawa, H. Unden, and L. Graeve. 2006. Identification of a
basolateral sorting signal within the cytoplasmic domain of the interleukin-6
signal transducer gp130. Cell Signal. 18:1140-1146.
164
Dreteler, G.H., W. Wouters, G.P. Toorop, J.A. Jansen, and P.R. Saxena. 1991. Systemic
and regional hemodynamic effects of the 5-hydroxytryptamine1A receptor
agonists flesinoxan and 8-hydroxy-2(di-N-propylamino)tetralin in the conscious
rat. J Cardiovasc Pharmacol. 17:488-493.
Duffy, J.B. 2002. GAL4 system in Drosophila: a fly geneticist's Swiss army knife.
Genesis. 34:1-15.
Ebrey, T., and Y. Koutalos. 2001. Vertebrate photoreceptors. Prog Retin Eye Res. 20:49-
94.
Edeling, M.A., S. Sanker, T. Shima, P.K. Umasankar, S. Honing, H.Y. Kim, L.A.
Davidson, S.C. Watkins, M. Tsang, D.J. Owen, and L.M. Traub. 2009. Structural
requirements for PACSIN/Syndapin operation during zebrafish embryonic
notochord development. PLoS One. 4:e8150.
Egan, C., K. Herrick-Davis, and M. Teitler. 1998. Creation of a constitutively activated
state of the 5-HT2A receptor by site-directed mutagenesis: revelation of inverse
agonist activity of antagonists. Ann N Y Acad Sci. 861:136-139.
Erickson, J.T., G. Shafer, M.D. Rossetti, C.G. Wilson, and E.S. Deneris. 2007. Arrest of
5HT neuron differentiation delays respiratory maturation and impairs neonatal
homeostatic responses to environmental challenges. Respir Physiol Neurobiol.
159:85-101.
Etter, P.D., R. Narayanan, Z. Navratilova, C. Patel, D. Bohmann, H. Jasper, and M.
Ramaswami. 2005. Synaptic and genomic responses to JNK and AP-1 signaling
in Drosophila neurons. BMC Neurosci. 6:39.
165
Fain, G.L., H.R. Matthews, and M.C. Cornwall. 1996. Dark adaptation in vertebrate
photoreceptors. Trends Neurosci. 19:502-507.
Fischer, M., S. Kaech, U. Wagner, H. Brinkhaus, and A. Matus. 2000. Glutamate
receptors regulate actin-based plasticity in dendritic spines. Nat Neurosci. 3:887-
894.
Florea, L., G. Hartzell, Z. Zhang, G.M. Rubin, and W. Miller. 1998. A computer program
for aligning a cDNA sequence with a genomic DNA sequence. Genome Res.
8:967-974.
Fotiadis, D., B. Jastrzebska, A. Philippsen, D.J. Muller, K. Palczewski, and A. Engel.
2006. Structure of the rhodopsin dimer: a working model for G-protein-coupled
receptors. Curr Opin Struct Biol. 16:252-259.
Francis, P.T. 1996. Pyramidal neurone modulation: a therapeutic target for Alzheimer's
disease. Neurodegeneration. 5:461-465.
Frost, A., V.M. Unger, and P. De Camilli. 2009. The BAR domain superfamily:
membrane-molding macromolecules. Cell. 137:191-196.
Gao, G., C. McMahon, J. Chen, and Y.S. Rong. 2008. A powerful method combining
homologous recombination and site-specific recombination for targeted
mutagenesis in Drosophila. Proc Natl Acad Sci U S A. 105:13999-14004.
Goley, E.D., and M.D. Welch. 2006. The ARP2/3 complex: an actin nucleator comes of
age. Nat Rev Mol Cell Biol. 7:713-726.
Gonzalez-Fernandez, F. 2002. Evolution of the visual cycle: the role of retinoid-binding
proteins. J Endocrinol. 175:75-88.
166
Gordon, J.A., and R. Hen. 2004. The serotonergic system and anxiety. Neuromolecular
Med. 5:27-40.
Gramates, L.S., and V. Budnik. 1999. Assembly and maturation of the Drosophila larval
neuromuscular junction. Int Rev Neurobiol. 43:93-117.
Greenspan, R.J. 2004. Fly pushing : the theory and practice of Drosophila genetics. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. xiv, 191 p. pp.
Grove, M., G. Demyanenko, A. Echarri, P.A. Zipfel, M.E. Quiroz, R.M. Rodriguiz, M.
Playford, S.A. Martensen, M.R. Robinson, W.C. Wetsel, P.F. Maness, and A.M.
Pendergast. 2004. ABI2-deficient mice exhibit defective cell migration, aberrant
dendritic spine morphogenesis, and deficits in learning and memory. Mol Cell
Biol. 24:10905-10922.
Habermann, B. 2004. The BAR-domain family of proteins: a case of bending and
binding? EMBO Rep. 5:250-255.
Halbach, A., M. Morgelin, M. Baumgarten, M. Milbrandt, M. Paulsson, and M. Plomann.
2007. PACSIN 1 forms tetramers via its N-terminal F-BAR domain. Febs J.
274:773-782.
Hall, A. 1998. Rho GTPases and the actin cytoskeleton. Science. 279:509-514.
Hamill, O.P., A. Marty, E. Neher, B. Sakmann, and F.J. Sigworth. 1981. Improved patch-
clamp techniques for high-resolution current recording from cells and cell-free
membrane patches. Pflugers Arch. 391:85-100.
Han, M., and T.P. Sakmar. 2000. Assays for activation of recombinant expressed opsins
by all-trans-retinals. Methods Enzymol. 315:251-267.
167
Hanson, M.G., L.D. Milner, and L.T. Landmesser. 2008. Spontaneous rhythmic activity
in early chick spinal cord influences distinct motor axon pathfinding decisions.
Brain Res Rev. 57:77-85.
Heisler, L.K., H.M. Chu, T.J. Brennan, J.A. Danao, P. Bajwa, L.H. Parsons, and L.H.
Tecott. 1998. Elevated anxiety and antidepressant-like responses in serotonin 5-
HT1A receptor mutant mice. Proc Natl Acad Sci U S A. 95:15049-15054.
Hendricks, T.J., D.V. Fyodorov, L.J. Wegman, N.B. Lelutiu, E.A. Pehek, B. Yamamoto,
J. Silver, E.J. Weeber, J.D. Sweatt, and E.S. Deneris. 2003. Pet-1 ETS Gene Plays
a Critical Role in 5-HT Neuron Development and Is Required for Normal
Anxiety-like and Aggressive Behavior. Neuron. 37:233-247.
Herlitze, S., and L.T. Landmesser. 2007. New optical tools for controlling neuronal
activity. Curr Opin Neurobiol. 17:87-94.
Higgs, H.N., and T.D. Pollard. 2001. Regulation of actin filament network formation
through ARP2/3 complex: activation by a diverse array of proteins. Annu Rev
Biochem. 70:649-676.
Hjorth, S. 1993. Serotonin 5-HT1A autoreceptor blockade potentiates the ability of the 5-
HT reuptake inhibitor citalopram to increase nerve terminal output of 5-HT in
vivo: a microdialysis study. J Neurochem. 60:776-779.
Hjorth, S., H.J. Bengtsson, A. Kullberg, D. Carlzon, H. Peilot, and S.B. Auerbach. 2000.
Serotonin autoreceptor function and antidepressant drug action. J
Psychopharmacol. 14:177-185.
Hodges, M.R., G.J. Tattersall, M.B. Harris, S.D. McEvoy, D.N. Richerson, E.S. Deneris,
R.L. Johnson, Z.F. Chen, and G.B. Richerson. 2008. Defects in breathing and
168
thermoregulation in mice with near-complete absence of central serotonin
neurons. J Neurosci. 28:2495-2505.
Hotulainen, P., and C.C. Hoogenraad. 2010. Actin in dendritic spines: connecting
dynamics to function. J Cell Biol. 189:619-629.
Hoyer, D., J.P. Hannon, and G.R. Martin. 2002. Molecular, pharmacological and
functional diversity of 5-HT receptors. Pharmacol Biochem Behav. 71:533-554.
Hummel, T., and C. Klambt. 2008. P-element mutagenesis. Methods Mol Biol. 420:97-
117.
Itoh, T., K.S. Erdmann, A. Roux, B. Habermann, H. Werner, and P. De Camilli. 2005.
Dynamin and the actin cytoskeleton cooperatively regulate plasma membrane
invagination by BAR and F-BAR proteins. Dev Cell. 9:791-804.
Jarecki, J., and H. Keshishian. 1995. Role of neural activity during synaptogenesis in
Drosophila. J Neurosci. 15:8177-8190.
Jin, S., T.D. McKee, and D.D. Oprian. 2003. An improved rhodopsin/EGFP fusion
protein for use in the generation of transgenic Xenopus laevis. FEBS Lett.
542:142-146.
Johansen, J., M.E. Halpern, K.M. Johansen, and H. Keshishian. 1989. Stereotypic
morphology of glutamatergic synapses on identified muscle cells of Drosophila
larvae. J Neurosci. 9:710-725.
Jolimay, N., L. Franck, X. Langlois, M. Hamon, and M. Darmon. 2000. Dominant role of
the cytosolic C-terminal domain of the rat 5-HT1B receptor in axonal-apical
targeting. J Neurosci. 20:9111-9118.
169
Kaksonen, M., C.P. Toret, and D.G. Drubin. 2006. Harnessing actin dynamics for
clathrin-mediated endocytosis. Nat Rev Mol Cell Biol. 7:404-414.
Kalkman, H.O. 1995. RU 24969-induced locomotion in rats is mediated by 5-HT1A
receptors. Naunyn Schmiedebergs Arch Pharmacol. 352:583-584.
Kandel, E.R., J.H. Schwartz, and T.M. Jessell. 2000. Principles of neural science.
McGraw-Hill, Health Professions Division, New York. xli, 1414 p. pp.
Karnik, S.S., C. Gogonea, S. Patil, Y. Saad, and T. Takezako. 2003. Activation of G-
protein-coupled receptors: a common molecular mechanism. Trends Endocrinol
Metab. 14:431-437.
Kasai, K., K. Suga, T. Izumi, and K. Akagawa. 2008. Syntaxin 8 has two functionally
distinct di-leucine-based motifs. Cell Mol Biol Lett. 13:144-154.
Kefalov, V.J., M. Carter Cornwall, and R.K. Crouch. 1999. Occupancy of the
chromophore binding site of opsin activates visual transduction in rod
photoreceptors. J Gen Physiol. 113:491-503.
Keshishian, H., and Y.S. Kim. 2004. Orchestrating development and function: retrograde
BMP signaling in the Drosophila nervous system. Trends Neurosci. 27:143-147.
Kessels, M.M., and B. Qualmann. 2002. Syndapins integrate N-WASP in receptor-
mediated endocytosis. Embo J. 21:6083-6094.
Kessels, M.M., and B. Qualmann. 2004. The syndapin protein family: linking membrane
trafficking with the cytoskeleton. J Cell Sci. 117:3077-3086.
Kessels, M.M., and B. Qualmann. 2006. Syndapin oligomers interconnect the
machineries for endocytic vesicle formation and actin polymerization. J Biol
Chem. 281:13285-13299.
170
Khodosh, R., A. Augsburger, T.L. Schwarz, and P.A. Garrity. 2006. Bchs, a BEACH
domain protein, antagonizes Rab11 in synapse morphogenesis and other
developmental events. Development. 133:4655-4665.
Kia, H.K., M.C. Miquel, M.J. Brisorgueil, G. Daval, M. Riad, S. El Mestikawy, M.
Hamon, and D. Verge. 1996. Immunocytochemical localization of serotonin1A
receptors in the rat central nervous system. J Comp Neurol. 365:289-305.
Kim, J., T. Dittgen, A. Nimmerjahn, J. Waters, V. Pawlak, F. Helmchen, S. Schlesinger,
P.H. Seeburg, and P. Osten. 2004. Sindbis vector SINrep(nsP2S726): a tool for
rapid heterologous expression with attenuated cytotoxicity in neurons. J Neurosci
Methods. 133:81-90.
Kim, J.M., J. Hwa, P. Garriga, P.J. Reeves, U.L. RajBhandary, and H.G. Khorana. 2005.
Light-driven activation of beta 2-adrenergic receptor signaling by a chimeric
rhodopsin containing the beta 2-adrenergic receptor cytoplasmic loops.
Biochemistry. 44:2284-2292.
Kim, Y., J.Y. Sung, I. Ceglia, K.W. Lee, J.H. Ahn, J.M. Halford, A.M. Kim, S.P. Kwak,
J.B. Park, S. Ho Ryu, A. Schenck, B. Bardoni, J.D. Scott, A.C. Nairn, and P.
Greengard. 2006. Phosphorylation of WAVE1 regulates actin polymerization and
dendritic spine morphology. Nature. 442:814-817.
Kiselev, A., and S. Subramaniam. 1994. Activation and regeneration of rhodopsin in the
insect visual cycle. Science. 266:1369-1373.
Kiselev, A., and S. Subramaniam. 1996. Modulation of arrestin release in the light-driven
regeneration of Rh1 Drosophila rhodopsin. Biochemistry. 35:1848-1855.
171
Kobilka, B.K., T. Frielle, S. Collins, T. Yang-Feng, T.S. Kobilka, U. Francke, R.J.
Lefkowitz, and M.G. Caron. 1987. An intronless gene encoding a potential
member of the family of receptors coupled to guanine nucleotide regulatory
proteins. Nature. 329:75-79.
Kofuji, P., N. Davidson, and H.A. Lester. 1995. Evidence that neuronal G-protein-gated
inwardly rectifying K+ channels are activated by G beta gamma subunits and
function as heteromultimers. Proc Natl Acad Sci U S A. 92:6542-6546.
Koh, T.W., V.I. Korolchuk, Y.P. Wairkar, W. Jiao, E. Evergren, H. Pan, Y. Zhou, K.J.
Venken, O. Shupliakov, I.M. Robinson, C.J. O'Kane, and H.J. Bellen. 2007.
Eps15 and Dap160 control synaptic vesicle membrane retrieval and synapse
development. J Cell Biol. 178:309-322.
Koh, T.W., P. Verstreken, and H.J. Bellen. 2004. Dap160/intersectin acts as a stabilizing
scaffold required for synaptic development and vesicle endocytosis. Neuron.
43:193-205.
Kroeze, W.K., K. Kristiansen, and B.L. Roth. 2002. Molecular biology of serotonin
receptors structure and function at the molecular level. Curr Top Med Chem.
2:507-528.
Kumar, V., S.R. Alla, K.S. Krishnan, and M. Ramaswami. 2009a. Syndapin is
dispensable for synaptic vesicle endocytosis at the Drosophila larval
neuromuscular junction. Mol Cell Neurosci. 40:234-241.
Kumar, V., R. Fricke, D. Bhar, S. Reddy-Alla, K.S. Krishnan, S. Bogdan, and M.
Ramaswami. 2009b. Syndapin promotes formation of a postsynaptic membrane
system in Drosophila. Mol Biol Cell. 20:2254-2264.
172
Kuriu, T., A. Inoue, H. Bito, K. Sobue, and S. Okabe. 2006. Differential control of
postsynaptic density scaffolds via actin-dependent and -independent mechanisms.
J Neurosci. 26:7693-7706.
Kusserow, H., B. Davies, H. Hortnagl, I. Voigt, T. Stroh, B. Bert, D.R. Deng, H. Fink,
R.W. Veh, and F. Theuring. 2004. Reduced anxiety-related behaviour in
transgenic mice overexpressing serotonin 1A receptors. Brain Res Mol Brain Res.
129:104-116.
Lahey, T., M. Gorczyca, X.X. Jia, and V. Budnik. 1994. The Drosophila tumor
suppressor gene dlg is required for normal synaptic bouton structure. Neuron.
13:823-835.
Lamb, T.D., and E.N. Pugh, Jr. 2004. Dark adaptation and the retinoid cycle of vision.
Prog Retin Eye Res. 23:307-380.
Lanfumey, L., and M. Hamon. 2004. 5-HT1 receptors. Curr Drug Targets CNS Neurol
Disord. 3:1-10.
Lembo, P.M., and P.R. Albert. 1995. Multiple phosphorylation sites are required for
pathway-selective uncoupling of the 5-hydroxytryptamine1A receptor by protein
kinase C. Mol Pharmacol. 48:1024-1029.
Leone, M., A. Attanasio, D. Croci, A. Ferraris, D. D'Amico, L. Grazzi, A. Nespolo, and
G. Bussone. 1998. 5-HT1A receptor hypersensitivity in migraine is suggested by
the m-chlorophenylpiperazine test. Neuroreport. 9:2605-2608.
Lerch-Haner, J.K., D. Frierson, L.K. Crawford, S.G. Beck, and E.S. Deneris. 2008.
Serotonergic transcriptional programming determines maternal behavior and
offspring survival. Nat Neurosci. 11:1001-1003.
173
Letourneur, F., and R.D. Klausner. 1992. A novel di-leucine motif and a tyrosine-based
motif independently mediate lysosomal targeting and endocytosis of CD3 chains.
Cell. 69:1143-1157.
Li, H.C., E.Y. Li, L. Neumeier, L. Conforti, and M. Soleimani. 2007. Identification of a
novel signal in the cytoplasmic tail of the Na+:HCO3- cotransporter NBC1 that
mediates basolateral targeting. Am J Physiol Renal Physiol. 292:F1245-1255.
Li, X., D.V. Gutierrez, M.G. Hanson, J. Han, M.D. Mark, H. Chiel, P. Hegemann, L.T.
Landmesser, and S. Herlitze. 2005a. Fast noninvasive activation and inhibition of
neural and network activity by vertebrate rhodopsin and green algae
channelrhodopsin. Proc Natl Acad Sci U S A. 102:17816-17821.
Li, X., A. Hummer, J. Han, M. Xie, K. Melnik-Martinez, R.L. Moreno, M. Buck, M.D.
Mark, and S. Herlitze. 2005b. G protein beta2 subunit-derived peptides for
inhibition and induction of G protein pathways. Examination of voltage-gated
Ca2+ and G protein inwardly rectifying K+ channels. J Biol Chem. 280:23945-
23959.
Li, Y.C., W.Z. Bai, L. Zhou, L.K. Sun, and T. Hashikawa. 2010. Nonhomogeneous
distribution of filamentous actin in the presynaptic terminals on the spinal
motoneurons. J Comp Neurol. 518:3184-3192.
Li, Z., J. Zhuang, and D.W. Corson. 1999. Delivery of 9-Cis retinal to photoreceptors
from bovine serum albumin. Photochem Photobiol. 69:500-504.
Lloyd, T.E., P. Verstreken, E.J. Ostrin, A. Phillippi, O. Lichtarge, and H.J. Bellen. 2000.
A genome-wide search for synaptic vesicle cycle proteins in Drosophila. Neuron.
26:45-50.
174
Lucki, I. 1998. The spectrum of behaviors influenced by serotonin. Biol Psychiatry.
44:151-162.
Mackler, J.M., J.A. Drummond, C.A. Loewen, I.M. Robinson, and N.E. Reist. 2002. The
C(2)B Ca(2+)-binding motif of synaptotagmin is required for synaptic
transmission in vivo. Nature. 418:340-344.
Malmberg, A., and P.G. Strange. 2000. Site-directed mutations in the third intracellular
loop of the serotonin 5-HT(1A) receptor alter G protein coupling from G(i) to
G(s) in a ligand-dependent manner. J Neurochem. 75:1283-1293.
Mann, J.J., D.A. Brent, and V. Arango. 2001. The neurobiology and genetics of suicide
and attempted suicide: a focus on the serotonergic system.
Neuropsychopharmacology. 24:467-477.
Marie, B., S.T. Sweeney, K.E. Poskanzer, J. Roos, R.B. Kelly, and G.W. Davis. 2004.
Dap160/intersectin scaffolds the periactive zone to achieve high-fidelity
endocytosis and normal synaptic growth. Neuron. 43:207-219.
Mark, M.D., and S. Herlitze. 2000. G-protein mediated gating of inward-rectifier K+
channels. Eur J Biochem. 267:5830-5836.
Marques, G., H. Bao, T.E. Haerry, M.J. Shimell, P. Duchek, B. Zhang, and M.B.
O'Connor. 2002. The Drosophila BMP type II receptor Wishful Thinking
regulates neuromuscular synapse morphology and function. Neuron. 33:529-543.
Marques, G., and B. Zhang. 2006. Retrograde signaling that regulates synaptic
development and function at the Drosophila neuromuscular junction. Int Rev
Neurobiol. 75:267-285.
175
Marrus, S.B., S.L. Portman, M.J. Allen, K.G. Moffat, and A. DiAntonio. 2004.
Differential localization of glutamate receptor subunits at the Drosophila
neuromuscular junction. J Neurosci. 24:1406-1415.
Mason, A.K., B.E. Jacobs, and P.A. Welling. 2008. AP-2-dependent internalization of
potassium channel Kir2.3 is driven by a novel di-hydrophobic signal. J Biol
Chem. 283:5973-5984.
Massarwa, R., S. Carmon, B.Z. Shilo, and E.D. Schejter. 2007. WIP/WASp-based actin-
polymerization machinery is essential for myoblast fusion in Drosophila. Dev
Cell. 12:557-569.
Mathew, D., B. Ataman, J. Chen, Y. Zhang, S. Cumberledge, and V. Budnik. 2005.
Wingless signaling at synapses is through cleavage and nuclear import of receptor
DFrizzled2. Science. 310:1344-1347.
Matus, A. 2000. Actin-based plasticity in dendritic spines. Science. 290:754-758.
Mayorga, A.J., A. Dalvi, M.E. Page, S. Zimov-Levinson, R. Hen, and I. Lucki. 2001.
Antidepressant-like behavioral effects in 5-hydroxytryptamine(1A) and 5-
hydroxytryptamine(1B) receptor mutant mice. J Pharmacol Exp Ther. 298:1101-
1107.
McBee, J.K., K. Palczewski, W. Baehr, and D.R. Pepperberg. 2001. Confronting
complexity: the interlink of phototransduction and retinoid metabolism in the
vertebrate retina. Prog Retin Eye Res. 20:469-529.
McCabe, B.D., S. Hom, H. Aberle, R.D. Fetter, G. Marques, T.E. Haerry, H. Wan, M.B.
O'Connor, C.S. Goodman, and A.P. Haghighi. 2004. Highwire regulates
presynaptic BMP signaling essential for synaptic growth. Neuron. 41:891-905.
176
McCabe, B.D., G. Marques, A.P. Haghighi, R.D. Fetter, M.L. Crotty, T.E. Haerry, C.S.
Goodman, and M.B. O'Connor. 2003. The BMP homolog Gbb provides a
retrograde signal that regulates synaptic growth at the Drosophila neuromuscular
junction. Neuron. 39:241-254.
McMahon, H.T., and J.L. Gallop. 2005. Membrane curvature and mechanisms of
dynamic cell membrane remodelling. Nature. 438:590-596.
Melia, T.J., Jr., C.W. Cowan, J.K. Angleson, and T.G. Wensel. 1997. A comparison of
the efficiency of G protein activation by ligand-free and light-activated forms of
rhodopsin. Biophys J. 73:3182-3191.
Melyan, Z., E.E. Tarttelin, J. Bellingham, R.J. Lucas, and M.W. Hankins. 2005. Addition
of human melanopsin renders mammalian cells photoresponsive. Nature.
433:741-745.
Merilainen, J., V.P. Lehto, and V.M. Wasenius. 1997. FAP52, a novel, SH3 domain-
containing focal adhesion protein. J Biol Chem. 272:23278-23284.
Modregger, J., B. Ritter, B. Witter, M. Paulsson, and M. Plomann. 2000. All three
PACSIN isoforms bind to endocytic proteins and inhibit endocytosis. J Cell Sci.
113 Pt 24:4511-4521.
Molina, T.M., S.C. Torres, A. Flores, T. Hara, R. Hara, and L.J. Robles. 1992.
Immunocytochemical localization of retinal binding protein in the octopus retina:
a shuttle protein for 11-cis retinal. Exp Eye Res. 54:83-90.
Mooney, R.D., T.A. Crnko-Hoppenjans, M. Ke, C.A. Bennett-Clarke, R.D. Lane, N.L.
Chiaia, and R.W. Rhoades. 1998. Augmentation of serotonin in the developing
177
superior colliculus alters the normal development of the uncrossed retinotectal
projection. J Comp Neurol. 393:84-92.
Moritz, O.L., B.M. Tam, D.S. Papermaster, and T. Nakayama. 2001. A functional
rhodopsin-green fluorescent protein fusion protein localizes correctly in
transgenic Xenopus laevis retinal rods and is expressed in a time-dependent
pattern. J Biol Chem. 276:28242-28251.
Murakami, M., and T. Kouyama. 2008. Crystal structure of squid rhodopsin. Nature.
453:363-367.
Muraki, Y., A. Yamanaka, N. Tsujino, T.S. Kilduff, K. Goto, and T. Sakurai. 2004.
Serotonergic regulation of the orexin/hypocretin neurons through the 5-HT1A
receptor. J Neurosci. 24:7159-7166.
Nahm, M., S. Kim, S.K. Paik, M. Lee, S. Lee, Z.H. Lee, J. Kim, D. Lee, and Y.C. Bae.
2010. dCIP4 (Drosophila Cdc42-interacting protein 4) restrains synaptic growth
by inhibiting the secretion of the retrograde Glass bottom boat signal. J Neurosci.
30:8138-8150.
Nebigil, C.G., P. Hickel, N. Messaddeq, J.L. Vonesch, M.P. Douchet, L. Monassier, K.
Gyorgy, R. Matz, R. Andriantsitohaina, P. Manivet, J.M. Launay, and L.
Maroteaux. 2001. Ablation of serotonin 5-HT(2B) receptors in mice leads to
abnormal cardiac structure and function. Circulation. 103:2973-2979.
Nelson, R.J., and S. Chiavegatto. 2001. Molecular basis of aggression. Trends Neurosci.
24:713-719.
178
O'Connor-Giles, K.M., L.L. Ho, and B. Ganetzky. 2008. Nervous wreck interacts with
thickveins and the endocytic machinery to attenuate retrograde BMP signaling
during synaptic growth. Neuron. 58:507-518.
Obosi, L.A., R. Hen, D.J. Beadle, I. Bermudez, and L.A. King. 1997. Mutational analysis
of the mouse 5-HT7 receptor: importance of the third intracellular loop for
receptor-G-protein interaction. FEBS Lett. 412:321-324.
Okajima, T.I., D.R. Pepperberg, H. Ripps, B. Wiggert, and G.J. Chader. 1990.
Interphotoreceptor retinoid-binding protein promotes rhodopsin regeneration in
toad photoreceptors. Proc Natl Acad Sci U S A. 87:6907-6911.
Okamoto, K., T. Nagai, A. Miyawaki, and Y. Hayashi. 2004. Rapid and persistent
modulation of actin dynamics regulates postsynaptic reorganization underlying
bidirectional plasticity. Nat Neurosci. 7:1104-1112.
Oksenberg, D., S. Havlik, S.J. Peroutka, and A. Ashkenazi. 1995. The third intracellular
loop of the 5-hydroxytryptamine2A receptor determines effector coupling
specificity. J Neurochem. 64:1440-1447.
Ozawa, Y., and N. Okado. 2002. Alteration of serotonergic receptors in the brain stems of
human patients with respiratory disorders. Neuropediatrics. 33:142-149.
Packard, M., E.S. Koo, M. Gorczyca, J. Sharpe, S. Cumberledge, and V. Budnik. 2002.
The Drosophila Wnt, wingless, provides an essential signal for pre- and
postsynaptic differentiation. Cell. 111:319-330.
Palczewski, K., T. Kumasaka, T. Hori, C.A. Behnke, H. Motoshima, B.A. Fox, I. Le
Trong, D.C. Teller, T. Okada, R.E. Stenkamp, M. Yamamoto, and M. Miyano.
179
2000. Crystal structure of rhodopsin: A G protein-coupled receptor. Science.
289:739-745.
Panda, S., S.K. Nayak, B. Campo, J.R. Walker, J.B. Hogenesch, and T. Jegla. 2005.
Illumination of the melanopsin signaling pathway. Science. 307:600-604.
Pardo, C.A., and C.G. Eberhart. 2007. The neurobiology of autism. Brain Pathol. 17:434-
447.
Parks, C.L., P.S. Robinson, E. Sibille, T. Shenk, and M. Toth. 1998. Increased anxiety of
mice lacking the serotonin1A receptor. Proc Natl Acad Sci U S A. 95:10734-
10739.
Parsons, L.H., T.M. Kerr, and L.H. Tecott. 2001. 5-HT(1A) receptor mutant mice exhibit
enhanced tonic, stress-induced and fluoxetine-induced serotonergic
neurotransmission. J Neurochem. 77:607-617.
Patel, T.D., and F.C. Zhou. 2005. Ontogeny of 5-HT1A receptor expression in the
developing hippocampus. Brain Res Dev Brain Res. 157:42-57.
Paterson, D.S., F.L. Trachtenberg, E.G. Thompson, R.A. Belliveau, A.H. Beggs, R.
Darnall, A.E. Chadwick, H.F. Krous, and H.C. Kinney. 2006. Multiple
serotonergic brainstem abnormalities in sudden infant death syndrome. JAMA.
296:2124-2132.
Pauwels, P.J., A. Gouble, and T. Wurch. 1999. Activation of constitutive 5-
hydroxytryptamine(1B) receptor by a series of mutations in the BBXXB motif:
positioning of the third intracellular loop distal junction and its G(o)alpha protein
interactions. Biochem J. 343 Pt 2:435-442.
180
Pepperberg, D.R., T.L. Okajima, B. Wiggert, H. Ripps, R.K. Crouch, and G.J. Chader.
1993. Interphotoreceptor retinoid-binding protein (IRBP). Molecular biology and
physiological role in the visual cycle of rhodopsin. Mol Neurobiol. 7:61-85.
Perkins, B.D., J.M. Fadool, and J.E. Dowling. 2004. Photoreceptor structure and
development: analyses using GFP transgenes. Methods Cell Biol. 76:315-331.
Peter, B.J., H.M. Kent, I.G. Mills, Y. Vallis, P.J. Butler, P.R. Evans, and H.T. McMahon.
2004. BAR domains as sensors of membrane curvature: the amphiphysin BAR
structure. Science. 303:495-499.
Pielage, J., R.D. Fetter, and G.W. Davis. 2005. Presynaptic spectrin is essential for
synapse stabilization. Curr Biol. 15:918-928.
Pielage, J., R.D. Fetter, and G.W. Davis. 2006. A postsynaptic spectrin scaffold defines
active zone size, spacing, and efficacy at the Drosophila neuromuscular junction.
J Cell Biol. 175:491-503.
Plomann, M., R. Lange, G. Vopper, H. Cremer, U.A. Heinlein, S. Scheff, S.A. Baldwin,
M. Leitges, M. Cramer, M. Paulsson, and D. Barthels. 1998. PACSIN, a brain
protein that is upregulated upon differentiation into neuronal cells. Eur J
Biochem. 256:201-211.
Pollard, T.D., and C.C. Beltzner. 2002. Structure and function of the Arp2/3 complex.
Curr Opin Struct Biol. 12:768-774.
Prokop, A. 2006. Organization of the efferent system and structure of neuromuscular
junctions in Drosophila. Int Rev Neurobiol. 75:71-90.
181
Prokop, A., M. Landgraf, E. Rushton, K. Broadie, and M. Bate. 1996. Presynaptic
development at the Drosophila neuromuscular junction: assembly and localization
of presynaptic active zones. Neuron. 17:617-626.
Prokop, A., and I.A. Meinertzhagen. 2006. Development and structure of synaptic
contacts in Drosophila. Semin Cell Dev Biol. 17:20-30.
Qanbar, R., and M. Bouvier. 2003. Role of palmitoylation/depalmitoylation reactions in
G-protein-coupled receptor function. Pharmacol Ther. 97:1-33.
Qiu, X., T. Kumbalasiri, S.M. Carlson, K.Y. Wong, V. Krishna, I. Provencio, and D.M.
Berson. 2005. Induction of photosensitivity by heterologous expression of
melanopsin. Nature. 433:745-749.
Qualmann, B., and R.B. Kelly. 2000. Syndapin isoforms participate in receptor-mediated
endocytosis and actin organization. J Cell Biol. 148:1047-1062.
Qualmann, B., M.M. Kessels, and R.B. Kelly. 2000. Molecular links between
endocytosis and the actin cytoskeleton. J Cell Biol. 150:F111-116.
Qualmann, B., J. Roos, P.J. DiGregorio, and R.B. Kelly. 1999. Syndapin I, a synaptic
dynamin-binding protein that associates with the neural Wiskott-Aldrich
syndrome protein. Mol Biol Cell. 10:501-513.
Racz, B., and R.J. Weinberg. 2008. Organization of the Arp2/3 complex in hippocampal
spines. J Neurosci. 28:5654-5659.
Ramboz, S., R. Oosting, D.A. Amara, H.F. Kung, P. Blier, M. Mendelsohn, J.J. Mann, D.
Brunner, and R. Hen. 1998. Serotonin receptor 1A knockout: an animal model of
anxiety-related disorder. Proc Natl Acad Sci U S A. 95:14476-14481.
Rando, R.R. 1996. Polyenes and vision. Chem Biol. 3:255-262.
182
Rao, Y., Q. Ma, A. Vahedi-Faridi, A. Sundborger, A. Pechstein, D. Puchkov, L. Luo, O.
Shupliakov, W. Saenger, and V. Haucke. 2010. Molecular basis for SH3 domain
regulation of F-BAR-mediated membrane deformation. Proc Natl Acad Sci U S A.
107:8213-8218.
Rawson, J.M., M. Lee, E.L. Kennedy, and S.B. Selleck. 2003. Drosophila neuromuscular
synapse assembly and function require the TGF-beta type I receptor saxophone
and the transcription factor Mad. J Neurobiol. 55:134-150.
Raymond, J.R., Y.V. Mukhin, A. Gelasco, J. Turner, G. Collinsworth, T.W. Gettys, J.S.
Grewal, and M.N. Garnovskaya. 2001. Multiplicity of mechanisms of serotonin
receptor signal transduction. Pharmacol Ther. 92:179-212.
Razzaq, A., I.M. Robinson, H.T. McMahon, J.N. Skepper, Y. Su, A.C. Zelhof, A.P.
Jackson, N.J. Gay, and C.J. O'Kane. 2001. Amphiphysin is necessary for
organization of the excitation-contraction coupling machinery of muscles, but not
for synaptic vesicle endocytosis in Drosophila. Genes Dev. 15:2967-2979.
Reiter, L.T., L. Potocki, S. Chien, M. Gribskov, and E. Bier. 2001. A systematic analysis
of human disease-associated gene sequences in Drosophila melanogaster. Genome
Res. 11:1114-1125.
Riad, M., S. Garcia, K.C. Watkins, N. Jodoin, E. Doucet, X. Langlois, S. el Mestikawy,
M. Hamon, and L. Descarries. 2000. Somatodendritic localization of 5-HT1A and
preterminal axonal localization of 5-HT1B serotonin receptors in adult rat brain. J
Comp Neurol. 417:181-194.
Richardson-Jones, J.W., C.P. Craige, B.P. Guiard, A. Stephen, K.L. Metzger, H.F. Kung,
A.M. Gardier, A. Dranovsky, D.J. David, S.G. Beck, R. Hen, and E.D. Leonardo.
183
2010. 5-HT1A autoreceptor levels determine vulnerability to stress and response
to antidepressants. Neuron. 65:40-52.
Richer, M., R. Hen, and P. Blier. 2002. Modification of serotonin neuron properties in
mice lacking 5-HT1A receptors. Eur J Pharmacol. 435:195-203.
Richerson, G.B. 2004. Serotonergic neurons as carbon dioxide sensors that maintain pH
homeostasis. Nat Rev Neurosci. 5:449-461.
Ritter, B., J. Modregger, M. Paulsson, and M. Plomann. 1999. PACSIN 2, a novel
member of the PACSIN family of cytoplasmic adapter proteins. FEBS Lett.
454:356-362.
Rodal, A.A., R.N. Motola-Barnes, and J.T. Littleton. 2008. Nervous wreck and Cdc42
cooperate to regulate endocytic actin assembly during synaptic growth. J
Neurosci. 28:8316-8325.
Rohatgi, R., L. Ma, H. Miki, M. Lopez, T. Kirchhausen, T. Takenawa, and M.W.
Kirschner. 1999. The interaction between N-WASP and the Arp2/3 complex links
Cdc42-dependent signals to actin assembly. Cell. 97:221-231.
Roy, S., and C.M. Hart. 2010. Targeted gene replacement by homologous recombination
in Drosophila stimulates production of second-site mutations. Fly (Austin). 4:12-
17.
Rubenstein, J.L. 1998. Development of serotonergic neurons and their projections. Biol
Psychiatry. 44:145-150.
Ruiz-Canada, C., and V. Budnik. 2006. Synaptic cytoskeleton at the neuromuscular
junction. Int Rev Neurobiol. 75:217-236.
184
Saari, J.C., G.G. Garwin, J.P. Van Hooser, and K. Palczewski. 1998. Reduction of all-
trans-retinal limits regeneration of visual pigment in mice. Vision Res. 38:1325-
1333.
Santarelli, L., M. Saxe, C. Gross, A. Surget, F. Battaglia, S. Dulawa, N. Weisstaub, J.
Lee, R. Duman, O. Arancio, C. Belzung, and R. Hen. 2003. Requirement of
hippocampal neurogenesis for the behavioral effects of antidepressants. Science.
301:805-809.
Sanyal, S., D.J. Sandstrom, C.A. Hoeffer, and M. Ramaswami. 2002. AP-1 functions
upstream of CREB to control synaptic plasticity in Drosophila. Nature. 416:870-
874.
Sari, Y. 2004. Serotonin1B receptors: from protein to physiological function and
behavior. Neurosci Biobehav Rev. 28:565-582.
Savelieva, K.V., S. Zhao, V.M. Pogorelov, I. Rajan, Q. Yang, E. Cullinan, and T.H.
Lanthorn. 2008. Genetic disruption of both tryptophan hydroxylase genes
dramatically reduces serotonin and affects behavior in models sensitive to
antidepressants. PLoS One. 3:e3301.
Schechter, L.E., R.H. Ring, C.E. Beyer, Z.A. Hughes, X. Khawaja, J.E. Malberg, and S.
Rosenzweig-Lipson. 2005. Innovative approaches for the development of
antidepressant drugs: current and future strategies. NeuroRx. 2:590-611.
Schmid, A., G. Qin, C. Wichmann, R.J. Kittel, S. Mertel, W. Fouquet, M. Schmidt, M.
Heckmann, and S.J. Sigrist. 2006. Non-NMDA-type glutamate receptors are
essential for maturation but not for initial assembly of synapses at Drosophila
neuromuscular junctions. J Neurosci. 26:11267-11277.
185
Schulein, R., R. Hermosilla, A. Oksche, M. Dehe, B. Wiesner, G. Krause, and W.
Rosenthal. 1998. A dileucine sequence and an upstream glutamate residue in the
intracellular carboxyl terminus of the vasopressin V2 receptor are essential for
cell surface transport in COS.M6 cells. Mol Pharmacol. 54:525-535.
Schuster, C.M., G.W. Davis, R.D. Fetter, and C.S. Goodman. 1996. Genetic dissection of
structural and functional components of synaptic plasticity. I. Fasciclin II controls
synaptic stabilization and growth. Neuron. 17:641-654.
Scott-McKean, J.J., G.R. Wenger, L.H. Tecott, and A.C. Costa. 2008. 5-HT(1A)
Receptor Null Mutant Mice Responding Under a Differential-Reinforcement-of-
Low-Rate 72-Second Schedule of Reinforcement. Open Neuropsychopharmacol
J. 1:24-32.
Scott, M.M., C.J. Wylie, J.K. Lerch, R. Murphy, K. Lobur, S. Herlitze, W. Jiang, R.A.
Conlon, B.W. Strowbridge, and E.S. Deneris. 2005. A genetic approach to access
serotonin neurons for in vivo and in vitro studies. Proc Natl Acad Sci U S A.
102:16472-16477.
Seletti, B., C. Benkelfat, P. Blier, L. Annable, F. Gilbert, and C. de Montigny. 1995.
Serotonin1A receptor activation by flesinoxan in humans. Body temperature and
neuroendocrine responses. Neuropsychopharmacology. 13:93-104.
Shaner, N.C., R.E. Campbell, P.A. Steinbach, B.N. Giepmans, A.E. Palmer, and R.Y.
Tsien. 2004. Improved monomeric red, orange and yellow fluorescent proteins
derived from Discosoma sp. red fluorescent protein. Nat Biotechnol. 22:1567-
1572.
186
Shapiro, D.A., K. Kristiansen, D.M. Weiner, W.K. Kroeze, and B.L. Roth. 2002.
Evidence for a model of agonist-induced activation of 5-hydroxytryptamine 2A
serotonin receptors that involves the disruption of a strong ionic interaction
between helices 3 and 6. J Biol Chem. 277:11441-11449.
Shimada, A., H. Niwa, K. Tsujita, S. Suetsugu, K. Nitta, K. Hanawa-Suetsugu, R.
Akasaka, Y. Nishino, M. Toyama, L. Chen, Z.J. Liu, B.C. Wang, M. Yamamoto,
T. Terada, A. Miyazawa, A. Tanaka, S. Sugano, M. Shirouzu, K. Nagayama, T.
Takenawa, and S. Yokoyama. 2007. Curved EFC/F-BAR-Domain Dimers Are
Joined End to End into a Filament for Membrane Invagination in Endocytosis.
Cell. 129:761-772.
Shimada, A., K. Takano, M. Shirouzu, K. Hanawa-Suetsugu, T. Terada, K. Toyooka, T.
Umehara, M. Yamamoto, S. Yokoyama, and S. Suetsugu. 2010. Mapping of the
basic amino-acid residues responsible for tubulation and cellular protrusion by the
EFC/F-BAR domain of pacsin2/Syndapin II. FEBS Lett. 584:1111-1118.
Shupliakov, O., P. Low, D. Grabs, H. Gad, H. Chen, C. David, K. Takei, P. De Camilli,
and L. Brodin. 1997. Synaptic vesicle endocytosis impaired by disruption of
dynamin-SH3 domain interactions. Science. 276:259-263.
Sigrist, S.J., D.F. Reiff, P.R. Thiel, J.R. Steinert, and C.M. Schuster. 2003. Experience-
dependent strengthening of Drosophila neuromuscular junctions. J Neurosci.
23:6546-6556.
Sotelo, C., B. Cholley, S. El Mestikawy, H. Gozlan, and M. Hamon. 1990. Direct
Immunohistochemical Evidence of the Existence of 5-HT1A Autoreceptors on
187
Serotoninergic Neurons in the Midbrain Raphe Nuclei. Eur J Neurosci. 2:1144-
1154.
Spang, A. 2004. Vesicle transport: a close collaboration of Rabs and effectors. Curr Biol.
14:R33-34.
Stamford, J.A., C. Davidson, D.P. McLaughlin, and S.E. Hopwood. 2000. Control of
dorsal raphe 5-HT function by multiple 5-HT(1) autoreceptors: parallel purposes
or pointless plurality? Trends Neurosci. 23:459-465.
Steru, L., R. Chermat, B. Thierry, and P. Simon. 1985. The tail suspension test: a new
method for screening antidepressants in mice. Psychopharmacology (Berl).
85:367-370.
Sweeney, S.T., and G.W. Davis. 2002. Unrestricted synaptic growth in spinster-a late
endosomal protein implicated in TGF-beta-mediated synaptic growth regulation.
Neuron. 36:403-416.
Sweitzer, S.M., and J.E. Hinshaw. 1998. Dynamin undergoes a GTP-dependent
conformational change causing vesiculation. Cell. 93:1021-1029.
Teller, D.C., T. Okada, C.A. Behnke, K. Palczewski, and R.E. Stenkamp. 2001.
Advances in determination of a high-resolution three-dimensional structure of
rhodopsin, a model of G-protein-coupled receptors (GPCRs). Biochemistry.
40:7761-7772.
Terakita, A., R. Hara, and T. Hara. 1989. Retinal-binding protein as a shuttle for retinal in
the rhodopsin-retinochrome system of the squid visual cells. Vision Res. 29:639-
652.
188
Terakita, A., T. Yamashita, S. Tachibanaki, and Y. Shichida. 1998. Selective activation
of G-protein subtypes by vertebrate and invertebrate rhodopsins. FEBS Lett.
439:110-114.
Thibault, S.T., M.A. Singer, W.Y. Miyazaki, B. Milash, N.A. Dompe, C.M. Singh, R.
Buchholz, M. Demsky, R. Fawcett, H.L. Francis-Lang, L. Ryner, L.M. Cheung,
A. Chong, C. Erickson, W.W. Fisher, K. Greer, S.R. Hartouni, E. Howie, L.
Jakkula, D. Joo, K. Killpack, A. Laufer, J. Mazzotta, R.D. Smith, L.M. Stevens,
C. Stuber, L.R. Tan, R. Ventura, A. Woo, I. Zakrajsek, L. Zhao, F. Chen, C.
Swimmer, C. Kopczynski, G. Duyk, M.L. Winberg, and J. Margolis. 2004. A
complementary transposon tool kit for Drosophila melanogaster using P and
piggyBac. Nat Genet. 36:283-287.
Thompson, J.D., D.G. Higgins, and T.J. Gibson. 1994. CLUSTAL W: improving the
sensitivity of progressive multiple sequence alignment through sequence
weighting, position-specific gap penalties and weight matrix choice. Nucleic
Acids Res. 22:4673-4680. van Roessel, P., D.A. Elliott, I.M. Robinson, A. Prokop, and A.H. Brand. 2004.
Independent regulation of synaptic size and activity by the anaphase-promoting
complex. Cell. 119:707-718.
Verstreken, P., T.W. Koh, K.L. Schulze, R.G. Zhai, P.R. Hiesinger, Y. Zhou, S.Q. Mehta,
Y. Cao, J. Roos, and H.J. Bellen. 2003. Synaptojanin is recruited by endophilin to
promote synaptic vesicle uncoating. Neuron. 40:733-748.
Vickers, S.P., and C.T. Dourish. 2004. Serotonin receptor ligands and the treatment of
obesity. Curr Opin Investig Drugs. 5:377-388.
189
Wan, H.I., A. DiAntonio, R.D. Fetter, K. Bergstrom, R. Strauss, and C.S. Goodman.
2000. Highwire regulates synaptic growth in Drosophila. Neuron. 26:313-329.
Wang, Q., M.V. Navarro, G. Peng, E. Molinelli, S.L. Goh, B.L. Judson, K.R.
Rajashankar, and H. Sondermann. 2009. Molecular mechanism of membrane
constriction and tubulation mediated by the F-BAR protein Pacsin/Syndapin. Proc
Natl Acad Sci U S A. 106:12700-12705.
Wegner, A.M., C.A. Nebhan, L. Hu, D. Majumdar, K.M. Meier, A.M. Weaver, and D.J.
Webb. 2008. N-wasp and the arp2/3 complex are critical regulators of actin in the
development of dendritic spines and synapses. J Biol Chem. 283:15912-15920.
Weiss, S., M. Sebben, D.E. Kemp, and J. Bockaert. 1986. Serotonin 5-HT1 receptors
mediate inhibition of cyclic AMP production in neurons. Eur J Pharmacol.
120:227-230.
Wess, J. 1997. G-protein-coupled receptors: molecular mechanisms involved in receptor
activation and selectivity of G-protein recognition. Faseb J. 11:346-354.
Wittemann, S., M.D. Mark, J. Rettig, and S. Herlitze. 2000. Synaptic localization and
presynaptic function of calcium channel beta 4-subunits in cultured hippocampal
neurons. J Biol Chem. 275:37807-37814.
Wu, C., R.W. Daniels, and A. DiAntonio. 2007. DFsn collaborates with Highwire to
down-regulate the Wallenda/DLK kinase and restrain synaptic terminal growth.
Neural Dev. 2:16.
Xia, Z., J.A. Gray, B.A. Compton-Toth, and B.L. Roth. 2003a. A direct interaction of
PSD-95 with 5-HT2A serotonin receptors regulates receptor trafficking and signal
transduction. J Biol Chem. 278:21901-21908.
190
Xia, Z., S.J. Hufeisen, J.A. Gray, and B.L. Roth. 2003b. The PDZ-binding domain is
essential for the dendritic targeting of 5-HT2A serotonin receptors in cortical
pyramidal neurons in vitro. Neuroscience. 122:907-920.
Xie, M., X. Li, J. Han, D.L. Vogt, S. Wittemann, M.D. Mark, and S. Herlitze. 2007.
Facilitation versus depression in cultured hippocampal neurons determined by
targeting of Ca2+ channel Cavbeta4 versus Cavbeta2 subunits to synaptic
terminals. J Cell Biol. 178:489-502.
Yuste, R., and T. Bonhoeffer. 2004. Genesis of dendritic spines: insights from
ultrastructural and imaging studies. Nat Rev Neurosci. 5:24-34.
Zelhof, A.C., and R.W. Hardy. 2004. WASp is required for the correct temporal
morphogenesis of rhabdomere microvilli. J Cell Biol. 164:417-426.
Zemelman, B.V., G.A. Lee, M. Ng, and G. Miesenbock. 2002. Selective photostimulation
of genetically chARGed neurons. Neuron. 33:15-22.
Zhao, Z.Q., M. Scott, S. Chiechio, J.S. Wang, K.J. Renner, R.W.t. Gereau, R.L. Johnson,
E.S. Deneris, and Z.F. Chen. 2006. Lmx1b is required for maintenance of central
serotonergic neurons and mice lacking central serotonergic system exhibit normal
locomotor activity. J Neurosci. 26:12781-12788.
Zhou, F.C., T.D. Patel, D. Swartz, Y. Xu, and M.R. Kelley. 1999. Production and
characterization of an anti-serotonin 1A receptor antibody which detects
functional 5-HT1A binding sites. Brain Res Mol Brain Res. 69:186-201.
Zito, K., D. Parnas, R.D. Fetter, E.Y. Isacoff, and C.S. Goodman. 1999. Watching a
synapse grow: noninvasive confocal imaging of synaptic growth in Drosophila.
Neuron. 22:719-729.
191
Zito, K., and K. Svoboda. 2002. Activity-dependent synaptogenesis in the adult
Mammalian cortex. Neuron. 35:1015-1017.
192