novel optogenetic approach reveals a function of cgmp in ... · ca - cornu ammonis cam - calmodulin...
TRANSCRIPT
Novel optogenetic approach reveals a function of cGMP in
synaptic plasticity and memory
by
Jelena Borovac
A thesis submitted in conformity with the requirements
for the degree of Doctor of Philosophy
Molecular Genetics
University of Toronto
© Copyright by Jelena Borovac 2019
ii
Novel optogenetic approach reveals a function of cGMP
in synaptic plasticity and memory
Jelena Borovac
Doctor of Philosophy
Molecular Genetics
University of Toronto
2019
Abstract
Synaptic plasticity is the activity-dependent structural and functional modulation of the
synapse, which is essential for learning and memory. These processes are regulated by
signaling pathways involving cAMP and cGMP nucleotide messengers. cAMP signaling is
associated with enhanced synaptic structure, function and memory at the hippocampus, but the
exact role of cGMP remains elusive. Here, I use a novel optogenetic approach to study the role
of cGMP in structural and functional plasticity in hippocampal neurons and memory in the
dentate gyrus (DG). Strong synaptic activation induces synapse enlargement called structural
long-term potentiation (sLTP), which is further enhanced by cAMP (cAMP-sLTP). By
blocking the cGMP/PKG pathway, I demonstrated that postsynaptic cGMP signaling
negatively regulates cAMP-sLTP. Using a two-photon optogenetic approach to activate a light-
sensitive guanylyl cyclase at targeted dendritic spines, I revealed that an increase in
iii
postsynaptic cGMP was sufficient to suppress the cAMP effect on structural potentiation, but
not induction of sLTP, suggesting a competitive bidirectional regulation of structural plasticity
by cAMP and cGMP. To address the role of cGMP in synaptic plasticity and memory, I adapted
the optogenetic technique to manipulate cGMP signaling in neurons of the DG, which is crucial
for memory. I demonstrated that postsynaptic cGMP signaling at the perforant path synapses
of the dentate gyrus (PP-DG) enhances functional synaptic potentiation. By utilizing a
comprehensive battery of tests for mouse behavior, I also revealed that cGMP signaling is
associated with enhanced anxiety and reference memory in mice. This project established a
novel optogenetic technique for targeted cGMP manipulation and provided the first direct look
at cGMP function from the level of the single synapse to the living mouse brain.
iv
Acknowledgments
I am grateful to Dr. Kenichi Okamoto for his patience and careful guidance throughout my
project. I was lucky to be trained by such a remarkable scientist and his lessons will continue
to serve me throughout my career.
I’d also like to thank my supervisory committee, Dr. Mei Zhen and Dr. Gabrielle Boulianne
for their support, kindness and encouragement, both scientific and personal.
I am thankful to Dr. Keizo Takao at the University of Toyama for his expert supervision of our
8-month collaboration and to his kind technician Asoyama-san for taking the time to help me
acculturate and enjoy living in Japan.
Thank you to all members of the Okamoto lab both past and present for personal and
professional support.
My greatest gratitude is reserved for my mom, Radojka, who moved to Canada so I can enjoy
opportunities like this one, and my sister Snježana, who is the best friend I could ever ask for.
v
Table of Contents Abstract ................................................................................................................................ ii
Acknowledgments .............................................................................................................. iv
Table of Contents ................................................................................................................. v
List of Figures ................................................................................................................... viii
Abbreviations ...................................................................................................................... xi
1 Introduction ..................................................................................................................... 1
1.1 Molecular mechanisms of chemical synapse structural plasticity ........................... 2
1.1.1 Dendritic spines ........................................................................................... 2
1.1.2 Structural synaptic plasticity of dendritic spines ......................................... 4
1.1.3 Molecular signaling in structural synaptic plasticity ................................... 6
1.1.4 cAMP and cGMP in structural synaptic plasticity .................................... 10
1.2 Molecular mechanisms of synapse functional plasticity ....................................... 13
1.2.1 Hippocampal neural circuit anatomy ......................................................... 13
1.2.2 Functional synaptic plasticity .................................................................... 15
1.2.3 Molecular mechanisms of PP-DG LTP ..................................................... 17
1.2.4 cAMP and cGMP in functional plasticity .................................................. 18
1.3 Molecular mechanisms of memory ....................................................................... 20
1.3.1 Testing episodic memory in mice .............................................................. 22
1.3.2 cAMP and cGMP in memory .................................................................... 25
1.4 Connection between plasticity and memory .......................................................... 26
Chapter 1: The function of cGMP in bi-directional structural synaptic plasticity ............ 27
1.5 Induction of bi-directional structural plasticity in dendritic spines ....................... 30
1.5.1 cAMP enhances structural potentiation of dendritic spines (cAMP- sLTP) ................................................................................................................... 34
vi
1.5.2 Negative regulation of cAMP- sLTP at the dendritic spine ...................... 37
1.6 Role of cGMP signaling in negative regulation of cAMP-sLTP ........................... 40
1.6.1 cGMP-PKG function in negative regulation of cAMP-sLTP ................... 42
1.7 Postsynaptic cGMP is sufficient to suppress cAMP-sLTP ................................... 45
1.7.1 Photoactivation properties of BlgC in vitro ............................................... 46
1.7.2 Detection of BlgC photoactivation by FLIM probe in vitro ...................... 48
1.7.3 Detection of BlgC photoactivation in living neurons ................................ 50
1.7.4 Postsynaptic cGMP is sufficient to supress cAMP- dependent sLTP enhancement .............................................................................................. 52
1.8 Discussion .............................................................................................................. 55
Chapter 2: The role of cGMP in functional plasticity and mouse behavior ...................... 59
2.1 The role of postsynaptic cGMP in functional plasticity ........................................ 62
2.1.1 Validation of BlgC expression and light-dependent activation in DG neurons ....................................................................................................... 63
2.1.2 Effect of postsynaptic cGMP on PP-DG L-LTP ....................................... 66
2.2 The role of cGMP in mouse behavior ................................................................... 72
2.2.1 BlgC expression and light-dependent activation in the mouse brain ........ 73
2.2.2 Effect of postsynaptic cGMP on mouse behavior ..................................... 75
2.3 Discussion .............................................................................................................. 94
Conclusion and Future Directions ..................................................................................... 97
Materials and Methods ................................................................................................ 100
3.1 DNA constructs ................................................................................................... 100
3.2 Animal care .......................................................................................................... 100
3.3 Two-photon imaging and uncaging ..................................................................... 100
3.3.1 Structural potentiation ............................................................................. 101
3.3.2 Depotentiation of cAMP-dependent structural enlargement ................... 102
vii
3.3.3 Two-photon optogenetic cGMP production ............................................ 102
3.4 BlgC photoactivation ........................................................................................... 103
3.4.1 cGMP detection by ELISA in vitro ......................................................... 103
3.4.2 cGMP detection by FLIM ........................................................................ 104
3.5 Pharmacology ...................................................................................................... 105
3.6 AAV packaging ................................................................................................... 105
3.7 Surgery: Stereotaxic microinjection and LED implant ....................................... 105
3.8 Fluorescence imaging of mouse hippocampal slices ........................................... 106
3.9 Electrophysiology ................................................................................................ 106
3.10 Statistical analysis ................................................................................................ 107
3.11 Mouse behavior ................................................................................................... 108
3.11.1 General Health and Neurological Screen (GHNS) .................................. 108
3.11.2 Light/Dark test (LD) ................................................................................ 109
3.11.3 Elevated Plus maze (EP) .......................................................................... 109
3.11.4 Open Field test (OF) ................................................................................ 110
3.11.5 Crawley’s Social Interaction test (CSI) ................................................... 110
3.11.6 Social Interaction test (SI) ....................................................................... 111
3.11.7 Object Localization test (OLT) ................................................................ 111
3.11.8 Barnes Maze (BM) .................................................................................. 112
3.11.9 T-Maze (Forced Alternation) (TMFA) .................................................... 113
References ....................................................................................................................... 115
viii
List of Figures Figure 1: Dendritic spine structure ............................................................................................ 3
Figure 2: Structural plasticity of dendritic spines ...................................................................... 5
Figure 3: Postsynaptic actin/CaMKIIβ regulation during structural potentiation ..................... 9
Figure 4: Postsynaptic cAMP and cGMP signaling in the dendritic spine ............................. 12
Figure 5: Anatomy of hippocampal neural circuits ................................................................. 14
Figure 6: Categories of memory .............................................................................................. 21
Figure 7: Memory tests used to study episodic memory in mice ............................................ 23
Figure 8: Caged glutamate uncaging using a two-photon microscope .................................... 32
Figure 9: Two-photon uncaging targeting individual dendritic spine ..................................... 33
Figure 10: Induction of structural potentiation in dendritic spines (sLTP) ............................. 35
Figure 11: cAMP enhances structural potentiation of dendritic spine (sLTP) ........................ 36
Figure 12: Direct effect of low frequency stimulation (LFS) on dendritic spine structure ..... 38
Figure 13: Suppression of the dendritic spine structure following cAMP-sLTP .................... 39
Figure 14: Signaling pathways guiding bi-directional structural plasticity ............................. 41
Figure 15: cGMP-dependent protein kinase (PKG) is required for suppression of structural
enlargement ............................................................................................................................. 43
Figure 16: Summary of the effects of specific cGMP pathway inhibitors .............................. 44
Figure 17: Characterization of BlgC activity in vitro. ............................................................. 47
Figure 18: FRET/FLIM detection of cGMP produced by BlgC photoactivation in vitro ....... 49
ix
Figure 19: Detection of cGMP produced by BlgC in living neurons ...................................... 51
Figure 20: Postsynaptic cGMP-dependent suppression of cAMP-sLTP ................................. 53
Figure 21: Summary of the postsynaptic cGMP effects on cAMP-dependent sLTP .............. 54
Figure 22: Supplementary bidirectional regulation of structural plasticity ............................. 57
Figure 23: Proposed model of cAMP-cGMP signaling cross-talk during bidirectional
plasticity ................................................................................................................................... 58
Figure 24: Schematic strategy of BlgC expression targeting the DG neurons of the mouse
brain ......................................................................................................................................... 64
Figure 25: Detection of cGMP produced by BlgC photo-activation in hippocampal slices ... 65
Figure 26: Basal synaptic properties of wild-type and BlgC expressing mice ........................ 67
Figure 27. Photoactivation of BlgC enhances functional long-term potentiation (L-LTP) at
PP-DG synapses ...................................................................................................................... 69
Figure 28: The effect of BlgC photoactivation on paired pulse ratio before and after tetanic
stimulation ............................................................................................................................... 71
Figure 29: Schematic of light-dependent cGMP signal delivery in the mouse brain .............. 74
Figure 30: Summary schematic of mouse behavior analysis ................................................... 76
Figure 31: General Health and Neurological Screening (GHNS) ........................................... 78
Figure 32: Light-Dark (LD) test .............................................................................................. 80
Figure 33: Elevated Plus Maze (EP) test ................................................................................. 81
Figure 34: Open field (OF) test ............................................................................................... 82
Figure 35: Crawley’s social interaction (CSI) test .................................................................. 84
Figure 36: Single chamber social interaction (SI) test ............................................................. 85
x
Figure 37: Object Location test (OLT) .................................................................................... 88
Figure 38: Barnes Maze (BM) training trials .......................................................................... 89
Figure 39: Barnes Maze (BM) probe test (PT1) ...................................................................... 90
Figure 40: Forced Alternation T-maze test (TMFA). .............................................................. 91
Table1: Summary of the mouse behavior testing battery ........................................................ 93
xi
Abbreviations
AAV - adeno-associated virus
AC - adenylyl cyclase
ACSF - artificial cerebrospinal fluid
ADF - actin depolymerizing factor
AID - autoinhibitory domain
AMPAR - alpha-amino-3-hydroxy-5-
methyl-4-isoxazole propionic acid receptors
Arp2/3 - actin-related protein 2/3 complex
BlgC - bacterial light-activated guanylyl
cyclase
BM - Barnes maze test
CA - Cornu ammonis
CaM - calmodulin
CaMKII - calcium-calmodulin dependent
protein kinase II
cAMP - cyclic adenosine monophosphate
cGMP - cyclic guanosine monophosphate
CNG - cyclic nucleotide gated ion channel
CREB - cAMP-responsive element binding
CSI - Crawley’s social interaction test
DG - dentate gyrus
EC - entorhinal cortex
ELISA - enzyme linked immunosorbent
assay
E-LTP - early long-term potentiation
EP - elevated plus maze test
F-actin - filamentous actin
EPSP - excitatory postsynaptic potentials
FLIM - fluorescence lifetime imaging
microscopy
FRET - Förster-resonance energy transfer
GABA - gamma-amino butyric acid
GC - guanylyl cyclase
GFP - green fluorescent protein
GHNS - general health and neurological
screen
HEK 293 - human embryonic kidney 293
HFS - high frequency stimulation
HP - hot plate test
IEG - immediate early genes
I/O - input/output
LD - light-dark test
xii
LFS - low-frequency stimulation
LFU - low-frequency uncaging
L-LTP - late long-term potentiation
LTD - long-term depression
LTP - long-term potentiation
MF - mossy fiber
mGluR - metabotropic glutamate receptor
MPP - medial perforant pathway
MTL - medial temporal lobe
NMDAR - n-methyl-d-aspartate receptor
NO - nitric oxide
NOS - nitric oxide synthase
OF - open field test
OLT - object location test
PDE - phosphodiesterase
PKA - protein kinase A
PKG - protein kinase G
PP - perforant path
PP1 - phosphatase 1
PP2A - phosphatase 2A
PPI - pre-pulse inhibition
PPR - paired-pulse depression
PSD - postsynaptic density
PT1 - probe test 1
RR - rota-rod test
SC - Schaffer collateral
SI - social interaction test
sLTP - structural long-term potentiation
sLTD - structural long-term depression
TM - T-maze test
TMFA - T-maze forces alternation
TMLR - T-maze left right discrimination
VASP - vasodilator-stimulated
phosphoprotein
1
1 Introduction
Components of this chapter have been published in the Molecular and Cell Neuroscience:
Borovac, J., Bosch, M., and Okamoto, K. (2018). Regulation of actin dynamics during
structural plasticity of dendritic spines: Signaling messengers and actin-binding proteins. Mol
Cell Neurosci. 91, 122-130.
A link to the published paper can be found at:
https://www.sciencedirect.com/science/article/pii/S1044743117304177?via%3Dihub
2
1.1 Molecular mechanisms of chemical synapse structural plasticity
1.1.1 Dendritic spines
Signal transmission across the synapse is the source of our movement, pain, addiction, memory
and the root of our existence. The chemical synapse is the site of communication between two
neurons, where the pre-synaptic axon containing neurotransmitter vesicles meets the post-
synaptic “dendritic spine” of another neuron [1, 2] (Figure 1).
Spine-like structures are investigated using various models, including C. elegans [3, 4],
Drosophila [5], honeybees [6], jewel fish [7], songbirds [8] and rodents [9]. Dendritic spines
were first discovered by Ramón y Cajal, and originally described as “thorns” because of the
way they protrude from the dendrite body [10] (Figure 1). These small dendritic units serve to
compartmentalize signaling at excitatory synapses throughout the central nervous system,
including the hippocampus [9].
Hippocampal dendritic spines reach an average size of 1µm3 [11], but their morphology ranges
in size and shape depending on spine maturity (Figure 1). For example, thin elongated spines,
called “filipodia”, are newly formed and usually transient [12], while large “mushroom”
shaped spines [13] represent the mature form (Figure 1). The dendritic spine structure is highly
dynamic and tightly correlated with synaptic strength and function [14].
3
Figure 1: Dendritic spine structure
(A) Schematic of a hippocampal Cornu Ammonis 1 (CA1) pyramidal neuron, forming a
synapse with the axon of a Cornu Ammonis 3 (CA3) pyramidal neuron, with the synapse
formation detailed in the subset.
(B) Classification of dendritic spine structure including filipodia, stubby, mushroom and thin
spines.
4
1.1.2 Structural synaptic plasticity of dendritic spines
“Structural synaptic plasticity” refers to the change in dendritic spine structure that occurs in
response to repeated synaptic activity, such as sensory stimulation or learning a new memory
[15, 16]. The role of structural plasticity is not well understood, but it is thought to provide a
physical trace that represents the path of synaptic activity [16].
Activity-dependent structural change was first observed in the sensory neuron synapses of the
marine mollusk Aplysia californica [17]. In the mammalian brain, structural plasticity is most
intensively investigated at the dendritic spines of hippocampal Cornu Ammonis 1 (CA1)
pyramidal neurons. Early evidence of electrically induced synaptic morphological change was
reported in the mid-1970s using electron microscopy [18]. However, structural plasticity
occurs quickly (in the order of seconds) and it is difficult to predict which spines will undergo
structural change in response to electric stimulation of whole neurons [19]. Therefore, different
approaches are used today to study structural plasticity at individual dendritic spines.
Development of two-photon microscopy in combination with fluorescent proteins has enabled
us to directly monitor the alterations of dendritic spine structure in living brain tissue [20].
Induction of synaptic potentiation (LTP) leads to an increase in synaptic size (structural LTP:
sLTP), while synaptic depression (LTD) causes its shrinkage (structural LTD: sLTD) [21, 22]
(Figure 2). The size of the dendritic spine and strength of synaptic function are tightly
correlated, suggesting that structural plasticity is an essential process in synaptic plasticity.
5
Figure 2: Structural plasticity of dendritic spines
Schematic of dendritic spine enlargement and shrinkage associated with synaptic potentiation
and depression.
6
1.1.3 Molecular signaling in structural synaptic plasticity
1.1.3.1 Role of actin as a major cytoskeletal protein in dendritic spines
Actin is the most abundant protein in eukaryotic cells. Among its three main isoforms (α, β and
γ), actin-β and γ are the two isoforms found in non-muscle cells such as neurons [23]. It is
highly enriched in dendritic spines, where it functions as the major cytoskeletal component that
forms and maintains dendritic spine morphology. Actin is also the main anchoring site for
many postsynaptic proteins, including N-Methyl-D-Aspartate Receptors (NMDARs) and
Alpha-Amino-3-Hydroxy-5-Methyl-4-Isoxazole Propionic Acid Receptors (AMPARs) [24-
28]. Studies using two-photon FRET (Förster resonance energy transfer) live imaging showed
that actin exists in a dynamic equilibrium between globular (G)-actin and filamentous (F)-actin
[22, 29]. Activity-dependent actin polymerization and depolymerization during synaptic
plasticity cause structural enlargement and shrinkage of dendritic spines. Honkura and
colleagues have demonstrated the presence of different pools of F-actin in the spine: (1) a
dynamic F-actin pool with a fast turnover at the periphery of the spine, (2) a stable F-actin pool
with slow turnover at the base of the spine and (3) an “enlargement” pool with slow turnover
that is formed in response to sLTP induction [30].
7
Regulation of postsynaptic actin during structural plasticity mediates rapid morphological
changes at the dendritic spine
F-actin assembly: Upon synaptic stimulation, actin within the dendritic spine is rapidly
polymerized (~ 1 min) into F-actin and the spine volume increases as a result [22]. The actin
binding protein Arp2/3 is also rapidly translocated to the potentiated spines [31]. Arp2/3 has
nucleation activity, which increases the complexity of actin filaments through branching [32,
33].
F-actin stabilization: Actin-binding proteins such as calcium-calmodulin dependent protein
kinase II β (CaMKIIβ), Drebrin, and α-actinin serve as stabilizers of the actin cytoskeleton by
crosslinking F-actin in bundles or linking F-actin to postsynaptic density (PSD) proteins [34-
38]. These three actin binding proteins are highly enriched in dendritic spines, and their
suppression affects spine formation and morphology [36, 39, 40].
F-actin disassembly: Cofilin-1 and the related actin depolymerizing factor (ADF) are expressed
in neurons and found in the dendritic spine [41]. Cofilin exerts a bidirectional effect on F-actin,
depending on its relative concentration to actin. At high concentrations, cofilin can bind to
actin filaments and promote their stabilization or even the nucleation of new filaments [42]. At
low concentrations, cofilin promotes F-actin disassembly by cutting the actin filaments
(severing) or by facilitating the removal of the actin monomers (depolymerization) [43].
8
1.1.3.2 Role of Ca2+ and CaMKII in structural synaptic potentiation (sLTP)
Synaptic potentiation causes Ca2+ influx into dendritic spines, which activates a series of
biochemical processes including CaMKII (Figure 3). CaMKII α and β are particularly
abundant and constitute 2% of the total protein in the rodent hippocampus [44]. CaMKIIα is
proposed to function as a molecular trigger for LTP, while CaMKIIβ binds to actin filaments
and regulates sLTP induction and maintenance [36, 45-47].
Regulation of postsynaptic CaMKIIβ during structural plasticity
CaMKII autophosphorylation/activation: CaMKIIβ binds to F-actin in its basal state, creating
a bundling or stabilizing effect [36], while also limiting the binding of actin modulator
molecules [45]. Upon synaptic stimulation, this interaction is abolished by Ca2+/CaM binding
[48] and the resulting auto-phosphorylation of CaMKII leads to detachment and release of F-
actin, thereby allowing access to actin modifying proteins to subsequently modify the structure.
CaMKII dephosphorylation/inactivation: PP1 phosphatase plays a crucial role in gating
CaMKII autophosphorylation through de-phosphorylation of Thr286 of CaMKII [49, 50] along
with phosphatase 2A (PP2A) which predominantly de-phosphorylates the soluble cytoplasmic
CaMKII [51]. The balance between phosphorylation and de-phosphorylation of CaMKII is an
essential regulator of synaptic and structural plasticity, as well as learning and memory.
9
Figure 3: Postsynaptic actin/CaMKIIβ regulation during structural potentiation
Schematic representation of the changes in dendritic spine in response to synaptic potentiation
(1) Actin bundling by CaMKII (calcium/calmodulin dependent protein kinase II) under
baseline conditions. (2) CaMKII activation and detachment from actin following potentiation.
(3) Actin polymerization and dendritic spine enlargement (4) CaMKII-actin re-bundling and
stabilization of the newly formed structure.
10
1.1.4 cAMP and cGMP in structural synaptic plasticity
Cyclic AMP (cAMP) is a ubiquitous second messenger with a variety of cellular functions
[52]. In dendritic spines, cAMP is synthesized by adenylyl cyclase (AC) in response to the
activation of receptors such as NMDAR, mGluR and dopamine receptors [53-57]. The
postsynaptic cAMP pathway plays an essential role in protein synthesis-dependent LTP (L-
LTP) [58-60], through the activation of cAMP-Responsive Element Binding (CREB)
transcription factor [61, 62]. Pharmacological application of the potent AC activator forskolin
during synaptic potentiation demonstrated that cAMP enhances structural synaptic potentiation
(sLTP) by further increasing dendritic spine enlargement (cAMP-sLTP) [63, 64].
Our recent results indicate that cAMP carries out this effect on structural potentiation through
a mechanism that is rapid and protein synthesis-independent [65]. Strong synaptic stimulation
triggers cAMP and cAMP-dependent protein kinase (PKA) in addition to the activation of
CaMKII [65]. cAMP/PKA signalling prolongs the time of CaMKIIβ activation during sLTP
by blocking protein phosphatase 1 (PP1) (which normally inactivates CaMKII by
dephosphorylation [66, 67]). Thus, the postsynaptic cAMP/PKA/PP1/CaMKIIβ pathway
regulates structural synaptic potentiation by extending the time window for actin cytoskeleton
remodelling (Figure 4).
Cyclic GMP (cGMP) is synthesized in response to activation of various receptors including
NMDARs, cholinergic and dopamine receptors [68-70]. Postsynaptic Ca2+ influx stimulates
nitric oxide synthase (nNOS) to produce nitric oxide (NO), which activates guanylyl cyclase
(GC) and leads to the synthesis of cGMP [71-73] and activation of its major effectors, including
protein kinase G (PKG) [74], phosphodiesterase 2 (PDE2) [75], and cyclic nucleotide gated
11
ion channel (CNG) [76]. cAMP and cGMP are both degraded by phosphodiesterases (PDE),
including cAMP-specific PDE4, cGMP-specific PDE5 and PDE2 which can hydrolyze both
[77, 78].
The role of cGMP in NMDAR-dependent structural potentiation of spines is not known.
However, some molecules that regulate spine structure are found downstream of the cGMP
pathway. For example, the activity-dependent formation of excitatory spines is regulated by
NO/cGMP/PKG-dependent phosphorylation of vasodilator-stimulated phosphoprotein
(VASP) [79, 80], a protein that is involved in the regulation of actin filaments [81]. The cGMP-
PKG pathway also causes phosphorylation-dependent inhibition of RhoA, which is a Rho
GTPase that stimulates actin polymerization via ROCK LIM-kinase-cofilin pathway [80, 82,
83]. These findings suggest that cGMP signaling has the capacity to directly affect synaptic
structural plasticity through its interaction with actin modifying proteins (Figure 4).
12
Figure 4: Postsynaptic cAMP and cGMP signaling in the dendritic spine
Schematic representation of cAMP signaling in the dendritic spine, showing ATP conversion
into cAMP by adenylyl cyclase (AC) enzyme, degradation by phosphodiesterase (PDE) and
the activation of protein kinase A (PKA). PKA phosphoregulates proteins in the dendritic
spine, including phosphatase 1 (PP1), which dephosphorylates calmodulin-dependent protein
kinase II (CaMKII). cGMP is produced by GTP conversion into cGMP by guanylyl cyclase
(GC) and degraded by phosphodiesterases (PDE2 and PDE5). cGMP activates protein kinase
G (PKG), which phosphoregulates proteins involved in modulating synaptic structure, such as
RhoA and vasodilator-stimulated phosphoprotein (VASP).
13
1.2 Molecular mechanisms of synapse functional plasticity
1.2.1 Hippocampal neural circuit anatomy
The hippocampus consists of the dentate gyrus (DG), which contains granule cells, and the
Cornu Ammonis areas (CA1, CA2, CA3), which contains pyramidal cells [84] (Figure 5).
The entorhinal cortex (EC) connects the hippocampus with nearly all other association cortices
and represents the main source of incoming/sensory information [85]. The EC sends excitatory
projections through the perforant path (PP) to granule cells of the DG, whose mossy fiber
projections activate CA3 pyramidal neurons, which in turn activate CA1 pyramidal cells
through the Schaffer collateral (SC) pathway [86]. This is the trisynaptic pathway (sequential
projections from EC to DG, CA3, and then to CA1). Direct projections from EC to CA1 and
EC to CA3 are called monosynaptic pathways. CA1 neurons are the major output from the
hippocampus, projecting to various brain regions, such as the subiculum, perirhinal cortex,
prefrontal cortex, and amygdala [86].
14
Figure 5: Anatomy of hippocampal neural circuits
Schematic of hippocampal neural circuits including the monosynaptic pathways (direct
projections from EC to CA1 and EC to CA3) and the trisynaptic pathway (sequential
projections from EC to DG, CA3, and then to CA1).
15
1.2.2 Functional synaptic plasticity
Repeated stimulation alters synaptic strength or efficacy in the long-term, through a process
called functional or synaptic plasticity [87]. Synaptic plasticity is considered a major
molecular/cellular model of learning and memory [88].
Synaptic plasticity can be experimentally modeled in acute hippocampal slices using
electrophysiology, where the application of repetitive trains of electric stimuli induces
potentiation that is recorded as a change in field excitatory postsynaptic potentials (fEPSPs)
[89]. Synaptic plasticity can be investigated in a number of neural circuits of the hippocampus,
including CA3–CA3, CA3–CA1, perforant path (PP)-CA1, perforant path (PP)-dentate gyrus
(DG), perforant path (PP)-CA3 and mossy fiber-CA3 synapses [90].
Depending on the properties and frequency of electric stimulation, various types of plasticity
can be induced, such as early long-term potentiation (E-LTP), late long-term potentiation L-
LTP and long-term depression (LTD).
Long-term potentiation (LTP) is the activity-dependent enhancement of synaptic function,
expressed as early-LTP (E-LTP) and protein synthesis dependent late-LTP (L-LTP) [91, 92].
During LTP, strong synaptic activation leads to a robust Ca2+ increase, which in turn causes
activation of kinases such as CaMKII [93] and the recruitment of a-amino-3-hydroxy-5-
methyl-4-isoxazolepropionic acid receptors (AMPARs) to the membrane, resulting in a further
increase in synaptic transmission.
16
Regulation of CaMKII during synaptic potentiation
At baseline conditions, the central autoinhibitory domain (AID) of the CaMKII holoenzyme
interacts with its catalytic domain to maintain an inactive state. However, in stimulated
synapses, Ca2+ levels increase and Ca2+/CaM binding disrupts this autoinhibitory conformation.
This initiates autophosphorylation of Thr286 in the AID of CaMKII [94] and Thr305/306 in
the Ca2+/CaM binding domain [49], resulting in its autonomous, self-perpetuating activity and
the phosphorylation of proteins such as AMPAR. Thus, CaMKII increases synaptic efficacy
even after its original activation triggers LTP [95, 96].
17
1.2.3 Molecular mechanisms of PP-DG LTP
Perforant path (PP)-dentate gyrus (DG) synapses exhibit post-synaptic NMDA receptor-
dependent LTP [97]. PP-DG LTP is initiated by glutamate from the pre-synaptic terminal
binding to the receptors on the dendritic spine, resulting in Ca2+ ion influx/depolarization and
a selective increase in AMPA receptors causing increased synaptic transmission [88, 98]. The
early phase of PP-DG LTP is mediated by CaMKII, protein kinase C, and MAP kinase and it
is independent of protein synthesis and PKA activation [99, 100]. Conversely, the late phase
of LTP (L-LTP) is protein synthesis-dependent, and involves the cAMP/PKA pathway. cAMP
activates the cAMP-Responsive Element Binding (CREB) transcription factor and upregulates
transcription during L-LTP [101, 102].
Protein synthesis-dependent PP-DG L-LTP also induces expression of immediate early genes
(IEGs) such as c-fos, Zif268, Arc and Homer, which the Tonegawa lab elegantly adapted for
optogenetic tagging and manipulation of neurons involved in memory [101, 103]. These
experiments linked signaling in PP-DG neural pathways with memory processing and validated
PP-DG as a good model for studying molecular signalling in plasticity and memory [104-107].
18
1.2.4 cAMP and cGMP in functional plasticity
It is well established that cAMP mediates the late-phase of LTP (L-LTP), through activation
of the CREB transcription factor [108-110]. cAMP signaling was demonstrated in hippocampal
mossy fiber/CA3 [58], CA1/CA3 [59, 60] and medial perforant pathway (MPP) L-LTP [99].
Hippocampus-specific protein kinase G I (PKG-I) knockout mice exhibit impaired late-phase
potentiation, suggesting that cGMP signaling also mediates L-LTP [111]. Inhibiting guanylyl
cyclase blocks the induction of hippocampal CA3/CA1 LTP [112, 113] and perfusion of cGMP
analogs induces long-lasting potentiation [114, 115], strongly suggesting a functional role of
cGMP in enhancing LTP.
In contrast, cGMP-dependent PKG-I is required for long-term depression (LTD) in the
Purkinje cells of the cerebellum [116-118]. These findings suggest that cGMP modulates
synaptic plasticity in both directions, depending on the brain area, through different signaling
mechanisms such as NMDAR-dependent and G-protein coupled signaling [119]. Importantly,
cGMP signaling is known to facilitate neurotransmitter release from the pre-synaptic neuron
during potentiation [115, 120-122]. More specifically, post-synaptic receptor stimulation
causes Ca2+/calmodulin-dependent activation of nitric oxide synthase (NOs) and the synthesis
of NO, which then travels back to the presynaptic site by retrograde signaling and stimulates
presynaptic synthesis of cGMP.
At the presynaptic site, cGMP aids in the assembly of neurotransmitter-containing vesicles,
leading to an increase in transmitter release and enhanced LTP [121]. Therefore, it is important
to distinguish the pre- and post-synaptic contributions of cGMP to synaptic plasticity. One
study addressed postsynaptic cGMP function using cultured hippocampal neurons, where
cGMP-dependent protein kinase G II (PKG-II) was found to modulate AMPAR trafficking
19
[123]. However, current pharmacological and genetic approaches make it difficult to
specifically target postsynaptic pathways during hippocampal LTP. Therefore, more precise
methods will be necessary to reveal the exact postsynaptic function of cGMP in synaptic
plasticity.
20
1.3 Molecular mechanisms of memory
Memory is the internalization of relevant information (acquisition), followed by its storage
(consolidation), further processing (reconsolidation) and remembering (retrieval) [124]. The
main memory types are conscious “declarative” memory and implicit or unconscious
“procedural” memory. The medial temporal lobe (MTL) processes declarative memory by
utilizing structures in the hippocampal region (CA fields, dentate gyrus, and subicular
complex) and the adjacent perirhinal, entorhinal, and parahippocampal cortices [125]. Subjects
with medial temporal lobe damage fail to process new declarative memory, but can still access
information from the past, suggesting that memory eventually becomes independent of the
hippocampus [125]. This process occurs through “systems consolidation”, whereby permanent
memory becomes stored in synapses of the neocortex, instead of the hippocampus [126].
However, even this memory is subject to further modification and error because of a process
called reconsolidation, which makes it temporarily labile again [127, 128]. Declarative or
explicit memory is subcategorized into “semantic” memory of facts and “episodic”, sometimes
termed autobiographical, memory of what, when and where [129] (Figure 6). Episodic
memory is a “recently evolved, late-developing, and past-oriented” memory system [130].
Episodic-like memory can be experimentally evaluated in animal models by manipulating their
environment using objects, aversive stimuli or changes in spatial configuration [131]. When
testing memory using mice, an important distinction is made between “reference memory” and
“working memory” (Figure 6). Working memory is the short-term storage and processing of
information, such as immediate surroundings and cues available within the trial [132].
Reference memory is long-term, and representative of information acquired throughout a
number of training trials [133].
21
Figure 6: Categories of memory
Schematic representing main categories of memory with focus on hippocampal memory and
the methods for testing it in rodents.
22
1.3.1 Testing episodic memory in mice
Memory can be evaluated in mice using various tests, such as Contextual Fear Conditioning,
Maze variations (Morris Water maze, Radial arm maze, T-maze, Barnes maze) and Object Test
variations (object location, object discrimination) [131, 134, 135]. Because working memory
represents immediate processing, memory tasks with a changing goal location can be used to
evaluate working memory. An example of this type of test is the Forced Alternation T-maze,
where the mouse chooses the left or right arm of the maze, depending on the trial [133, 136].
Tests with a permanent goal location, such as the Barnes maze, can be used to evaluate
reference memory [136]. It is considered a reliable test of reference memory because mice are
trained over several trials to learn the permanent location of an escape hole [137]. Compared
to the Morris water maze, this test is advantageous because it removes the stress associated
with swimming, and offers a dry and spacious arena more compatible with the use of
optogenetics. The T-Maze test can be designed to evaluate either reference memory in the “T-
Maze Left/Right Discrimination”, or working memory using “T-Maze Forced Alternation”
[138]. The motivation in this test is a sucrose pellet reward, and some extra stress may arise
from the food restriction period preceding the test.
Finally, object localization tests commonly rely on short-term working memory because mice
are evaluated based on their spatial memory of the arena layout/components without training
trials [139]. I utilized these three tests in combination with optogenetics to study the role of
cGMP in memory (Figure 7).
23
Figure 7: Memory tests used to study episodic memory in mice
Schematic representing the memory tasks utilized to evaluate working memory (Object test
and T-Maze Forced Alternation) and reference memory (Barnes maze) in mice.
24
1.3.1.1 Memory in the hippocampal dentate gyrus (DG)
Selective lesions of the dentate gyrus (DG) cause impairment of spatial working memory, as
well as spatial reference memory [140-142], suggesting that the DG is a good model for
studying memory in mice. Recent channelrhodopsin optogenetic protocols specifically target
DG pathways to manipulate memory. Optogenetic activation of neurons in the DG was used
to (1) create a false memory in mice [143], (2) switch fear memory to reward memory through
re-association [144] and (3) acutely suppress depression-like behavior [145].
25
1.3.2 cAMP and cGMP in memory
Mutant mice lacking type I adenylyl cyclase (which generates cAMP) exhibit spatial memory
deficits in the Morris water maze test [146], while mice without functional cAMP-dependent
PKA show impaired memory retrieval [147]. Both of these results suggest that cAMP/PKA
signaling is crucial for memory. Indeed, a number of studies have confirmed the involvement
of cAMP in long-term memory and the potential of PKA as a therapeutic target for memory
disorders [148].
cGMP function in memory was addressed through pharmacological inhibition of
phosphodiesterase (PDE) enzymes, which degrade cGMP, cAMP or both [78]. PDE 2
hydrolyzes cAMP and cGMP, while PDE 5 and PDE 9 are cGMP-specific [149]. Remarkably,
PDE 2 [150], PDE 5 [151], and PDE 9 [152] inhibition have all been reported to significantly
improve learning and memory in rodents. This suggests that upregulation of cGMP signaling
is associated with improved memory. However, these findings do not rule out the possibility
of unspecific effects of PDE inhibition, such as elevated cAMP when PDE 2 is inhibited.
Furthermore, inhibitor of PDE 5 was reported to reverse cognitive dysfunction in Alzheimer’s
rodent models, indicating potential cGMP impairment in neurodegenerative diseases [153-
155].
26
1.4 Connection between plasticity and memory
Structural plasticity is considered by some to be the physical representation of synaptic
plasticity and the anatomical basis of memory formation [16, 156, 157]. A number of reports
have attempted to directly link morphological changes in dendritic spines with changes in
synaptic function [158, 159] and behavior [160].
Structural plasticity and memory: Early reports showed that the time course of neuronal
morphological change correlates with behavioral duration of memory [161], while more recent
studies also demonstrated that learning strongly correlates with enlargement of the post-
synaptic density [162]. Patients with cognitive disorders such as Alzheimer’s disease show
aberrant dendritic spine structure and their eventual loss [163, 164]. Together, these results
suggest a synapse-level structural dysfunction, and a tight correlation between synaptic
structure and memory [165].
Functional plasticity and memory: A number of reports have also attempted to demonstrate the
connection between synaptic potentiation (LTP) and memory. Reports showed that
hippocampus-dependent learning leads to LTP at hippocampal synapses, while suppression of
LTP after learning a task abolishes the memory of that task [166-168].
More research is necessary to determine the overlapping pathways between physical,
physiological and behavioral expressions of the changing synapse. By utilizing novel
optogenetic techniques to manipulate cGMP signaling at the level of the synapse, neuron and
the brain, I revealed a dynamic function of postsynaptic cGMP in (1) structural plasticity at a
targeted dendritic spine, (2) functional plasticity at the PP-DG synapse, and (3) memory in the
dentate gyrus of the hippocampus.
27
Chapter 1: The function of cGMP in bi-directional structural synaptic plasticity
In this Chapter, the FRET/FLIM probe characterization and cGMP visualization in vitro and
in neurons (Figure 18 and Figure 19) was performed with the kind assistance of my colleague
Dr. Tyler Luyben.
28
Chapter 1
Dendritic spines are the post-synaptic sites of excitatory synapses that undergo
structural change during synaptic plasticity. Induction of synaptic long-term potentiation (LTP)
leads to an enlargement in spine size (structural LTP: sLTP), whereas long-term depression
(LTD) induces spine shrinkage [21, 22, 169]. Actin is a major cytoskeletal protein in the
dendritic spine, which underlies structural change [22] through its activity-dependent
interaction with CaMKII. CaMKII β bundles actin filaments to stabilize the spine structure
[36]. During LTP induction, postsynaptic NMDAR-dependent influx of Ca2+ activates
CaMKIIα/β, which then detaches from actin filaments allowing for actin polymerization to
occur and change the shape of dendritic spines [22, 36, 170].
Intracellular messengers, such as cyclic AMP (cAMP) and cyclic GMP (cGMP) are also
involved in modulating plasticity [171]. Postsynaptic cAMP plays a key role in the protein
synthesis-dependent, late phase of long-term potentiation (L-LTP) [108-110, 172-175].
Increasing cAMP signaling during structural potentiation (sLTP) significantly enhances
dendritic spine enlargement (compared to sLTP without cAMP) [64].
Stimulation of NMDA receptors also activates nitric oxide (NO)-dependent cGMP
biosynthesis through calcium-calmodulin (Ca2+/CaM) stimulation of neuronal nitric oxide
synthase (nNOS) [69, 71, 72]. The resultant cGMP activates its major effectors, including
protein kinase G (PKG) [74], phosphodiesterase 2 (PDE2) [75] and cyclic nucleotide gated-
ion channel (CNG) [76]. In Purkinje cells of the cerebellum, knocking out cGMP-dependent
PKG-I leads to impaired LTD [116], suggesting that postsynaptic cGMP may be involved in
the negative regulation of synaptic potentiation.
In some systems such as axon/dendrite development in hippocampal neurons, cAMP and
cGMP signaling pathways cross-over and carry out reciprocal actions [176]. Although the role
29
of cAMP and cGMP in synaptic plasticity has been reported on, the role of postsynaptic cGMP
in relation to cAMP and structural plasticity is not yet known.
cAMP and cGMP signaling at the synapse is dynamic and it is difficult to study using
traditional pharmacological and genetic methods. These approaches cause global changes in
signaling, making it difficult to target the signaling pathways within individual dendritic
spines. To address this, I established an optogenetic microscopy technique using two-photon
excitation of a photoactivatable Bacterial Light-activated Guanylyl Cyclase (BlgC) [177] that
enabled me to spatiotemporally manipulate postsynaptic cGMP in living neurons.
Using this method, I identified the postsynaptic function of cGMP as a negative regulator of
cAMP-dependent structural enhancement of the dendritic spine. Competitive cGMP function
requires the protein kinase G (PKG) pathway and occurs independently of basal structural
potentiation. Therefore, I propose a dynamic postsynaptic interaction mechanism between
cAMP and cGMP that governs bi-directional structural plasticity through the effects on
kinase/phosphatase activity within dendritic spines.
30
1.5 Induction of bi-directional structural plasticity in dendritic spines
Targeting dendritic spines by two-photon caged glutamate uncaging
The studies of structural plasticity have expanded substantially since two-photon microscopy
was adapted for targeted synapse activation. Two-photon imaging can achieve localized
excitation at a focal point with an area as small as 1µm3 [20] and penetrate as deep as 1mm
into living tissue [178], making it a perfect tool for live imaging and localized photo-
manipulation. Precise two-photon stimulation can be used to photo-activate an MNI-caged
glutamate compound, converting the caged glutamate into its active form [179], which then
binds receptors to induce potentiation (Figure 8). The 1µm3 resolution of this technique
ensures that active glutamate only reaches the receptors of the targeted dendritic spine and
stimulates that spine (Figure 9).
Inducing structural plasticity
Strong synaptic activation induces structural potentiation (sLTP), which is accompanied by
enlargement of the dendritic spine. Matsuzaki and colleagues [158] were the first to use the
high-frequency two-photon uncaging protocol to demonstrate rapid enlargement of the
stimulated spines, which can last several hours [180]. Furthermore, it is known that structural
enlargement associated with sLTP can be further enhanced by cAMP (cAMP-sLTP) [63, 64].
Dendritic spine shrinkage is saturable, reversible, and requires NMDA receptor activation
[169, 181]. However, its discovery is more recent and it is much less understood compared to
structural enlargement. Previously, spine shrinkage was experimentally induced by low-
frequency electric stimulation (LFS, 1 Hz for 15 min) [182] and low-frequency glutamate
31
uncaging (LFU) in combination with postsynaptic depolarization [169]. These induction
protocols may recruit different signaling pathways within the dendritic spine, making it
difficult to draw mechanistic conclusions about structural shrinkage.
Using the established high frequency uncaging method for sLTP induction in combination with
a low frequency uncaging protocol, I demonstrated targeted bi-directional structural plasticity
and the signaling pathways involved in structural modulation of the dendritic spine.
32
Figure 8: Caged glutamate uncaging using a two-photon microscope
(A) In a conventional microscope (left), excitation light covers a broad area. In a two-photon
microscope (right), excitation is roughly double the wavelength of and therefore does not excite
photoactivatable molecules, such as proteins. As a result, the focal illumination area is as small
as 1µm3.
(B) Schematic structure of 4-Methoxy-7-nitroindolinyl-caged-L-glutamate (MNI-caged
glutamate) before and after uncaging by light. The “cage” component of the molecule is
cleaved by light and the “uncaged" glutamate serves as a neurotransmitter that activates
postsynaptic functions in excitatory neurons.
33
Figure 9: Two-photon uncaging targeting individual dendritic spine
Schematic of two-photon caged-glutamate uncaging at a target dendritic spine. Caged
glutamate is everywhere but it is uncaged (activated) only in the focal plane of the two-photon
laser (~1 μm), mimicking the synaptic activity-dependent release of glutamate from
presynaptic vesicles. Glutamate activates receptors on the surface of the postsynaptic dendritic
spine, which induces structural potentiation.
34
1.5.1 cAMP enhances structural potentiation of dendritic spines (cAMP- sLTP)
To induce and observe sLTP in organotypic rat hippocampal slices, I biolistically transfected
green fluorescent protein (GFP, as a volume filler) in CA1 pyramidal neurons and measured
the time-dependent structural change of dendritic spine size (3D volume) using a two-photon
fluorescence microscope.
I used a high frequency two-photon glutamate uncaging protocol (HFS) that induces structural
potentiation of dendritic spines [179]. The targeted dendritic spines were enlarged to
approximately double the size within 60 seconds, then and gradually decreased but maintained
in size (~125 % - sLTP) (Figure 10).
To visualize the effect of cAMP on sLTP, I increased cAMP levels by bath application of the
adenylate cyclase activator forskolin [183] (50µM, 5 min) in combination with sLTP induction
by caged glutamate uncaging. The induction of sLTP in the presence of forskolin enhanced
spine enlargement compared to control, which was maintained at the 30-minute time point
(Figure 11).
35
Figure 10: Induction of structural potentiation in dendritic spines (sLTP)
(A) Schematic of the HFS protocol for inducing sLTP. MNI-glutamate was added to ACSF
and activated by two-photon light near the targeted spine (~1 µm from the tip of the spine head)
using an established HFS uncaging protocol (HFS: 720nm, 1Hz, 4 msec duration, 10 mW, 60
sec).
(B) The images of dendritic spine enlargement were captured using two-photon GFP imaging
(900 nm excitation, 495-540 nm emission) and averaged to generate a time course of spine size
change over time. The spine size at baseline was set as 100 %. All error bars represent SEM.
(n = 21 dendritic spines/ 11 neurons; Exponential fit). Scale bar is 1μm.
36
Figure 11: cAMP enhances structural potentiation of dendritic spine (sLTP)
(A) Schematic and (B) Time course of cAMP-dependent sLTP. cAMP-dependent sLTP was
induced by high frequency caged glutamate uncaging (HFS) in the presence of the potent
adenylate cyclase activator forskolin (50µM, 5 min), represented with blue trace. The spine
size at baseline was set as 100 %. All error bars represent SEM. (n=17 dendritic spines/10
neurons; Exponential fit). Scale bar is 1μm.
37
1.5.2 Negative regulation of cAMP- sLTP at the dendritic spine
To visualize negative regulation of dendritic spine structure (structural depotentiation), I
adapted a two-photon low-frequency stimulation (LFS) uncaging protocol (15mW, 0.1Hz
paired pulsing with 50msec interval, no depolarization) [169]. When applied to naïve/basal
dendritic spines, LFS alone has no shrinkage effect on the dendritic spine structure (Figure
12).
Maximally enlarged dendritic spines (spines that are potentiated in the presence of cAMP) may
be more susceptible to negative regulation. To test whether the size of maximally enlarged
dendritic spines can be reduced by LFS, I induced cAMP-dependent sLTP (in the presence of
forskolin) immediately followed by LFS targeting the same dendritic spine.
The resulting structural plasticity was reduced to a size that is similar to spines which are
exposed to sLTP alone. These findings indicate that LFS eliminates (or suppresses) the cAMP-
dependent enhancement of the dendritic spine structure (Figure 13).
38
Figure 12: Direct effect of low frequency stimulation (LFS) on dendritic spine structure
(A) Schematic of low-frequency caged glutamate uncaging protocol (LFS: 0.1 Hz, paired
pulses with 0.5 msec duration of activation and 50 msec intervals, 15 mW power for 15
minutes).
(B) The time course of spine size change over time. The spine size at baseline was set as 100
%. All error bars represent SEM. (n = 12 dendritic spines/9 neurons, Linear fit). The horizontal
scale break represents the 15-minute low frequency stimulation period.
39
Figure 13: Suppression of the dendritic spine structure following cAMP-sLTP
(A) Schematic and (B) Time course of dendritic spine size change resulting from (1) sLTP
induced by HFS (black trace; Exponential fit), (2) sLTP induced by HFS in the presence of
forskolin (blue trace; Exponential fit) and (3) sLTP induced by HFS in the presence of forskolin
immediately followed by LFS (grey trace, n=25 dendritic spines/19 neurons; Exponential fit).
The spine size at baseline was set as 100 %. The horizontal scale break represents the 15-
minute low frequency stimulation period. Scale bar is 1μm. All error bars represent SEM.
40
1.6 Role of cGMP signaling in negative regulation of cAMP-sLTP
Postsynaptic cAMP signaling enhances structural potentiation of dendritic spines (cAMP-
sLTP) [65], but very little is known about the mechanisms involved in its negative regulation.
In some systems, such as calcium current in cardiac cells [184] and axon/dendrite formation in
cultured hippocampal neurons [176], cAMP and cGMP signaling mediate antagonistic or
reciprocal cellular actions. In undifferentiated neurites, Shelly and colleagues found that cAMP
promotes axon formation, while cGMP suppresses it [176]. Competition between cGMP and
cAMP signaling can occur at various stages, including cGMP-dependent activation of cAMP
hydrolysis by phosphodiesterases [185, 186] and competition for the same phosphorylation
substrate sites between PKG and PKA [187].
Since cGMP signaling is known to play a role in plasticity [188], and was previously implicated
in long-term depression (LTD) [117, 118] , I hypothesize that competitive regulation between
cAMP and cGMP governs bi-directional structural plasticity. More specifically, if cAMP
enhances structural enlargement, I predict that cGMP negatively regulates cAMP-dependent
structural enhancement (Figure 14).
41
Figure 14: Signaling pathways guiding bi-directional structural plasticity
Prediction of the signaling pathways involved in bi-directional structural plasticity. Since
cAMP enhances dendritic spine potentiation, cGMP signaling may be involved in negative
regulation of the dendritic spine enlargement (observed as a result of LFS).
42
1.6.1 cGMP-PKG function in negative regulation of cAMP-sLTP To test whether endogenous cGMP is directly involved in the negative regulation of cAMP-
dependent sLTP, I applied low frequency stimulation targeting cAMP-enlarged spines in the
presence of specific inhibitors of the major cGMP pathways (Figure 15). None of the bath-
applied inhibitors had a significant effect on the initial structural enlargement of the dendritic
spine (1 minute after cAMP-dependent sLTP) (Figure 15), suggesting that the cGMP pathway
is not involved in the initial phase of structural potentiation. In contrast, the bath-application
of PKG inhibitor KT5823 [189] abolished the LFS-dependent suppression of cAMP-dependent
sLTP (Figure 15). Other inhibitors such as PDE2 inhibitor (Bay 60-7550) [150] and CNG
inhibitor (L-cis-diltiazem) [76] (Figure 16) did not affect the suppression effect of LFS,
indicating the specific involvement of the cGMP/PKG pathway in negative regulation of
cAMP-sLTP.
43
Figure 15: cGMP-dependent protein kinase (PKG) is required for suppression of
structural enlargement
(A) Schematic representation and (B) time course of dendritic spine size change resulting from
(1) cAMP-dependent sLTP: sLTP induced by HFS in the presence of forskolin (n=17 dendritic
spines/10 neurons, blue trace; Exponential fit) (2) Suppression: sLTP induced by HFS in the
presence of forskolin immediately followed by LFS (grey trace, n=25 dendritic spines/19
neurons; Exponential fit) and (3) Blocking suppression: sLTP induced by HFS in the presence
of forskolin immediately followed by LFS in the presence of the PKG inhibitor KT 5823 (2µM,
n=15 dendritic spines/9 neurons; yellow trace; Exponential fit). The spine size at baseline was
set as 100 %. Horizontal scale break represents the 15-minute low frequency stimulation
period. Scale bar is 1μm. All error bars represent SEM.
44
Figure 16: Summary of the effects of specific cGMP pathway inhibitors
Average dendritic spine size at 30 minutes and 1 minute (inset) resulting from (1) cAMP-
dependent sLTP: sLTP induced by HFS in the presence of forskolin (n=17 dendritic spines/10
neurons, blue) (2) Suppression: sLTP induced by HFS in the presence of forskolin immediately
followed by LFS (n=25 dendritic spines/19 neurons, grey) and (3) Blocking suppression: sLTP
induced by HFS in the presence of forskolin immediately followed by LFS in the presence of
the PKG inhibitor KT 5823 (2µM, n=15 dendritic spines/9 neurons; yellow). Two other
effectors downstream of cGMP were inhibited in attempts to block suppression, including
PDE2 inhibitor Bay 60-7550 (100nM, n=16 dendritic spines/10 neurons, brown) and CNG
channel inhibitor L-cis-diltiazem (100µM, n=11 dendritic spines/7 neurons, green). In order to
compare all data against the sLTP + cAMP control, statistical analysis was performed using
Steel-Dunnett equivalent non-parametric test.
45
1.7 Postsynaptic cGMP is sufficient to suppress cAMP-sLTP
cGMP signaling can be increased by pharmacologically blocking its degradation using
inhibitors of specific phosphodiesterases, but the resulting cGMP signaling upregulation
cannot be timed, nor targeted to study post-synaptic function [150].
To determine whether post-synaptic cGMP is sufficient to depotentiate cAMP-dependent
sLTP, I established a two-photon optogenetic approach that can be used to increase cGMP
signaling within the same dendritic spine that is targeted for cAMP-dependent sLTP. Unlike
standard channelrhodopsin optogenetics that are used to photomanipulate neurons [190], this
approach utilizes a bacterial light activated guanylyl cyclase enzyme (BlgC), which locally
synthesizes cGMP when exposed to light [191]. Manipulation by light is less invasive
compared to certain pharmacological, genetic and electric stimulation approaches. With the
help of Dr. Tyler Luyben, who characterized a FRET/FLIM cGMP sensor, we confirmed that
BlgC photoactivation generates cGMP in living neurons and within targeted dendritic spines.
By specific manipulation of postsynaptic cGMP at targeted dendritic spines during sLTP, I
demonstrated that cGMP alone is sufficient to induce depotentiation of cAMP-dependent
sLTP.
46
1.7.1 Photoactivation properties of BlgC in vitro
To photo-manipulate cGMP levels, I utilized a blue-light activated guanylyl cyclase enzyme
(BlgC), which synthesizes cGMP when exposed to blue light [191]. Using blue LED (455 nm)
light, I photo-activated human codon optimized BlgC fused with red fluorescence protein
(tdTomato ([192]) and measured cGMP increase by cGMP ELISA in vitro (Figure 17). cGMP
levels increased within seconds of light exposure, confirming the photoactivation of BlgC.
I next measured two-photon wavelength-dependent BlgC photoactivation (Figure 17). The
small focal volume of two-photon excitation enables the manipulation of light-sensitive
proteins at targeted dendritic spines within deep brain tissues [45]. I found that shorter two-
photon excitation wavelengths (700 – 800 nm) efficiently activated BlgC in vitro. To avoid
possible photodamage of dendritic spines by shorter wavelengths, I photo-activated BlgC using
1000 nm light, which was previously successful in photo-manipulation of living neurons [45].
cGMP was synthesized following two-photon light stimulation (1000 nm), indicating
successful photoactivation of BlgC in vitro (Figure 17). In order to confirm the photoactivation
of BlgC to produce cGMP by 1000 nm light, we also used a cGMP FRET/FLIM (Förster-
resonance energy transfer/ fluorescence lifetime imaging microscopy) probe (cGiR), described
below.
47
Figure 17: Characterization of BlgC activity in vitro.
(A) Diagram of BlgC photoactivation by activation of blue light receptor using BLUF domain
and subsequent production of cGMP by guanylyl cyclase or GC domain (B) Schematic of LED
photoactivation of HEK 293 cell lysate expressing BlgC. (C) Time course of LED light-
dependent (455nm, 4.5 mW/mm2) cGMP production by BlgC. Following BlgC
photoactivation, cGMP concentration was measured in cell lysates using an ELISA. Data are
mean ± SEM (n = 3). Time course was fitted with a logistic curve. Inset: Off response of BlgC.
Production of cGMP in the absence of light (black bar) following 30 seconds of photoactivation
(white bar; n = 9; linear fit). (D) Schematic of two-photon photoactivation of HEK 293 cell
lysate expressing BlgC (1µl illumination area) (E) Two-photon excitation spectrum of BlgC
activity in vitro. cGMP concentration was measured in cell lysates expressing BlgC by an
ELISA following two-photon excitation (30 mW, 700 – 1,025 nm, 15 min) in vitro. Data are
mean ± SEM (n = 12). Cubic fitting, peak cGMP production using two-photon excitation was
defined as 100 %. (F) Time course of two-photon laser light-dependent (1,000 nm, 75 mW)
BlgC activation in vitro (n = 9; logistic fit). Data are mean ± SEM.
48
1.7.2 Detection of BlgC photoactivation by FLIM probe in vitro
To visualize photoactivation of BlgC in living neurons, we prepared a genetically-encoded
cGMP FRET/FLIM (Förster-resonance energy transfer/ fluorescence lifetime imaging
microscopy) probe (cGiR) utilizing a cGMP binding domain from PKG (Protein kinase G)
[193] fused with yellow (Citrine) and red fluorescence protein (tdTomato) (Figure 18). This
FRET/FLIM approach was necessary to avoid photobleaching of CFP during photoactivation
of BlgC. We first characterized the sensitivity of cGiR probe under a two-photon FRET/FLIM
microscope (900 nm excitation) and found a dose dependent response to cGMP measured as a
change in cGiR fluorescence lifetime (EC50 = 0.48 µM, Figure 18).
To detect light-dependent cGMP synthesis by BlgC using the FRET/FLIM probe in vitro, we
expressed the cGiR probe with and without BlgC in HEK293 cells and used the cell lysates to
assay fluorescence lifetime change in cGiR following photoactivation by focal light (414 ± 21
nm, mercury arc lamp) (Figure 18). We also directly measured cGMP concentration in the
same samples using a cGMP ELISA assay (Figure 18). Both the cGiR fluorescence lifetime
and the cGMP concentration were increased only in the presence of BlgC and focal light
stimulation, validating the use of the cGMP probe for visualization of BlgC photoactivation.
The FRET/FLIM experiment was performed with the kind help of my colleague Dr. Tyler
Luyben.
49
Figure 18: FRET/FLIM detection of cGMP produced by BlgC photoactivation in vitro
(A) Schematic representation of cGMP FRET probe activity. When cGMP binds to the cGIR
probe the resulting change in confirmation increases the distance between YFP and GFP,
therefore decreasing FRET and increasing fluorescence lifetime change (B) Dose-dependent
change in the fluorescence lifetime of cGiR probe in response to 8-Br-cGMP (n = 3). 8-Br-
cGMP was used to avoid degradation of cGMP. EC50 = 0.48 µM (Hill fit). (C) Schematic of
combined BlgC photoactivation (left) and cGMP detection (right) by FRET/FLIM probe
(cGiR). When BlgC is photoactivated, the cGMP produced binds to the cGIR probe which
leads to a measurable lifetime change. (D) Detection of BlgC photoactivation by fluorescence
lifetime change of cGiR (left) and cGMP ELISA (right). cGMP levels were measured in cell
lysates expressing cGiR with or without BlgC (BlgC +/-), and with or without light (light +/-).
Data are mean ± SEM. (Left) BlgC+/Light- (n = 10), BlgC-/Light+ (n = 15), and BlgC+/Light+
(n = 10). (Right) BlgC+/Light- (n = 3), BlgC-/Light+ (n = 3), and BlgC+/Light+ (n = 3).
50
1.7.3 Detection of BlgC photoactivation in living neurons
To validate photoactivation of BlgC and observe the cGMP synthesized in living neurons, we
expressed the cGiR cGMP probe with and without the BlgC in CA1 pyramidal neurons of rat
hippocampal organotypic cultured slices. Upon blue light stimulation (414 ± 21 nm),
fluorescence lifetime of cGMP probe (cGiR) was increased and maintained for at least 10
minutes only in the neurons co-expressing both BlgC and cGiR probe, indicating light-
dependent cGMP synthesis by BlgC in those neurons (Figure 19). We next targeted cGMP
synthesis to individual dendritic spines by two-photon excitation light (Figure 19). Following
two-photon photoactivation of BlgC at the targeted dendritic spines (1000 nm, 30 sec), the
fluorescence lifetime of the cGiR probe increased in the targeted spines and nearby dendrite,
then gradually retuned to baseline (τ = 4.61 min). Fluorescence lifetime in the absence of BlgC
was similar to baseline, indicating that cGiR activation is a result of cGMP generated by BlgC
photoactivation at the dendritic spines.
51
Figure 19: Detection of cGMP produced by BlgC in living neurons
(A) Schematic of BlgC photoactivation by focal light in CA1 pyramidal neurons of cultured
rat hippocampal slices. (B) Left: cGMP FRET/FLIM pseudo-colour images of a neuron before
and immediately after focal light photoactivation (414 ± 21 nm, 11 mW) of BlgC. Right: Time
course of fluorescence lifetime change of cGMP probe (cGiR) in neurons expressing cGiR
with or without BlgC (+BlgC/ -BlgC). +BlgC (n = 16), -BlgC (n = 11). (C) Schematic of BlgC
photoactivation in targeted dendritic spines using two-photon light (D) Left: cGMP
FRET/FLIM pseudo-colour images of a neuron before and immediately after BlgC two-photon
photoactivation (1000 nm, 11 mW, 30 sec) in the dendritic spine (blue circle). Right: Time-
course of the fluorescence lifetime change of cGMP probe (cGiR) in targeted dendritic spines
expressing cGiR with or without BlgC (+BlgC/ -BlgC). Exponential curve fitting (black dotted
line: τ = 4.61 min). +BlgC (n = 9), -BlgC (n = 10). Scale bar is 1μm.
52
1.7.4 Postsynaptic cGMP is sufficient to supress cAMP- dependent sLTP enhancement
Using two-photon BlgC activation at targeted dendritic spines, I next examined whether
postsynaptic cGMP is sufficient to suppress cAMP-dependent sLTP enhancement.
To test this, I biolistically co-expressed BlgC and GFP (as volume filler) in CA1 pyramidal
neurons of organotypic hippocampal slices and measured the time-dependent changes of
dendritic spine size during sLTP in the presence of cAMP (forskolin) and cGMP (two-photon
photoactivation of BlgC at target dendritic spines).
The light-dependent production of cGMP during sLTP alone did not affect dendritic spine
structure (Figure 20, inset), suggesting that cGMP does not affect basal structural potentiation.
I then tested the effect of postsynaptic cGMP synthesis on cAMP-dependent sLTP. Two-
photon activation of BlgC combined with cAMP-sLTP eliminated the cAMP-dependent
enhancement of structural potentiation (to the same level as basal sLTP), without affecting
initial spine enlargement (Figure 21). Furthermore, the bath-application of forskolin alone did
not affect spine structure (Figure 20, inset). These results indicate that cGMP signaling
negatively regulates cAMP-dependent enhancement of structural potentiation (sLTP) (Figure
21).
53
Figure 20: Postsynaptic cGMP-dependent suppression of cAMP-sLTP
(A) Schematic of two-photon activation of BlgC to produce cGMP in a targeted dendritic spine
(1000 nm, 11 mW, 60 sec) in combination with cAMP-dependent sLTP.
(B) Time course of spine size change over time resulting from (1) sLTP, induced HFS (n=21
dendritic spines/11 neurons, black trace; Exponential fit) (2) cAMP-dependent sLTP: induced
by HFS in the presence of forskolin (n=17 dendritic spines/10 neurons, blue trace; Exponential
fit) and (3) cGMP-dependent suppression: induced by BlgC photoactivation (1 minute,
1000nm) immediately followed by HFS in the presence of forskolin (n=17 dendritic spines/10
neurons, magenta; Exponential fit). Inset: (1) Direct effect of cGMP on sLTP induced by two-
photon activation of BlgC (60 sec, 1000 nm) in combination HFS in the absence of forskolin
(n = 10 dendritic spines/ 10 neurons, tan; Exponential fit) (2) Direct effect of cAMP on
dendritic spine structure induced by cAMP produced by forskolin (50 µM, 5 min), without
potentiation (n = 22 dendritic spines/ 10 neurons, light blue, Linear fit). The spine size at
baseline was set as 100 %. Scale bar is 1μm. All error bars represent SEM.
54
Figure 21: Summary of the postsynaptic cGMP effects on cAMP-dependent sLTP
Summary of spine size change at 30 min and 1 min (inset) following induction of (1) sLTP,
induced HFS (n=21 dendritic spines/11 neurons, black) (2) cAMP-dependent sLTP: induced
by HFS in the presence of forskolin (n=17 dendritic spines/10 neurons, blue) and (3) cGMP-
dependent suppression: induced by BlgC photoactivation (1 minute, 1000nm) immediately
followed by HFS in the presence of forskolin (n=17 dendritic spines/10 neurons, magenta) and
(4) cGMP-sLTP induced by two-photon photoactivation of BlgC (60 sec, 1000 nm)
immediately followed by induction of sLTP (HFS) in the absence of forskolin (n = 10 dendritic
spines/ 10 neurons, tan) Statistical significance compared with sLTP induction (black: HFS
only) by Steel test. All error bars represent SEM.
55
1.8 Discussion In this chapter, I demonstrated bi-directional structural plasticity at the dendritic spine
and showed that cGMP is required and sufficient for the negative regulation of cAMP-
dependent structural potentiation. Structural enlargement of living dendritic spines can be
experimentally induced by two-photon glutamate uncaging and further enhanced by cAMP
(cAMP-sLTP). To demonstrate negative regulation of the dendritic spine structure, I adapted
a low frequency glutamate uncaging protocol (LFS), which had no detectable effect on basal
dendritic spine structure. However, when applied to dendritic spines maximally enlarged by
cAMP-dependent potentiation, LFS of glutamate consistently abolished the cAMP-dependent
enhancement (to the same level as basal sLTP). To address the endogenous signaling
mechanisms, I induced LFS-dependent depotentiation of cAMP-sLTP in the presence of
specific inhibitors targeting the cGMP pathway. I found that inhibition of cGMP-dependent
protein kinase (PKG), but not phosphodiesterase 2 (PDE2) or the cyclic nucleotide gated
(CNG) channel, blocked LFS-dependent depotentiation, suggesting that endogenous
cGMP/PKG signaling is involved in negative regulation of cAMP-dependent sLTP. To test
whether increasing postsynaptic cGMP signaling is sufficient to negatively regulate cAMP-
dependent sLTP, I established a two-photon optogenetic technique for targeted cGMP photo-
manipulation. Using bacterial light sensitive guanylyl cyclase (BlgC), I showed that single-
photon and two-photon excitation light both rapidly synthesize cGMP in vitro. Furthermore,
we validated a cGMP FRET/FLIM sensor (cGiR) for detection of light-dependent cGMP by
two-photon FRET/FLIM microscopy and co-expressed it with BlgC in hippocampal slices to
measure the single- and two-photon light-dependent cGMP synthesis in living neurons. BlgC
photoactivation at targeted dendritic spines showed an instantaneous increase of cGMP,
56
indicating that BlgC is a suitable non-invasive tool that can be used to precisely manipulate
cGMP levels in deep brain tissues.
I increased cGMP levels by two-photon photoactivation of BlgC at targeted dendritic spines
immediately before induction of cAMP-dependent sLTP. Postsynaptic cGMP suppressed
cAMP-dependent enhancement of structural potentiation but did not affect basal
Ca2+/CaMKII-dependent sLTP, suggesting that the cGMP effect is reciprocal to cAMP and
independent of basal sLTP. In contrast, neither cGMP nor cAMP had a significant effect on
the initial spine enlargement. Since forskolin application to generate cAMP can enhance
structural potentiation (but is not necessary to induce it), I propose that the effect of cAMP is
supplementary to basal potentiation, and exists to enhance synaptic structure only under certain
conditions. This enhancement is negatively regulated by cGMP which, together with cAMP,
governs the supplementary bidirectional micromanipulation of the dendritic spine (Figure 22).
In the dendritic spine, cAMP/PKA and cGMP/PKG signaling can cross over through their
effects on protein phosphatase 1 (PP1) [194, 195], which indirectly controls CaMKII-
dependent synaptic potentiation [67]. Therefore, cAMP and cGMP may be competing for the
kinase/phosphatase-dependent regulation of PP1 activity in the dendritic spine in order to carry
out their effects on the synaptic structure [66] (Figure 23).
This bidirectional modulation of the synapse may play a crucial role in the processes of learning
and memory. Furthermore, our novel optogenetic protocol may provide an important
contribution to our understanding of the pathways involved in plasticity and memory.
57
Figure 22: Supplementary bidirectional regulation of structural plasticity
Schematic of the proposed supplementary bidirectional manipulation of the dendritic spine
structure by cAMP and cGMP
58
Figure 23: Proposed model of cAMP-cGMP signaling cross-talk during bidirectional
plasticity
Strong synaptic stimulation leads to postsynaptic activation of CaMKII by Ca2+ influx while
triggering the cAMP/PKA pathway at the same time. As PKA inactivates protein phosphatase
1 (PP1), which dephosphorylates and inactivates CaMKII, an increased presence of cAMP in
the spine may prolong the time of CaMKII activation during synaptic plasticity [67]
(enhancement, right) likely by extending the duration of actin cytoskeleton remodeling.
Alternately, NO-dependent synthesis of cGMP by sGC results in cGMP-dependent activation
of PKG, which may carry out its downstream effects through PP2A phosphatase-dependent
activation of PP1 and subsequent inactivation of CaMKIIβ (suppression, left). In this model,
cGMP leads to suppression of active CaMKII /actin polymerization, directly opposing the
action of cAMP through the effects on PP1 phosphatase.
59
Chapter 2: The role of cGMP in functional plasticity and mouse behavior
In this Chapter, the electrophysiology experiments were performed in collaboration with Dr.
John Roder and Dr. Graham Collingridge laboratories, using shared facilities and training
kindly provided to me by Dr. John Georgiou.
The mouse behaviour experiments were performed in collaboration with Dr. Keizo Takao and
Dr. Kaoru Inokuchi at the University of Toyama, Japan. These experiments were supported by
an International Collaboration award I received from the Weston Brain Institute in 2016, which
funded my 8-month stay in Japan.
60
Chapter 2
In the first chapter, I characterized a novel optogenetic method for targeted cGMP
signaling and demonstrated a role of the cGMP-PKG pathway in the negative regulation of
cAMP-dependent sLTP. A previous report from our lab showed that blocking structural
potentiation also prevents functional LTP [45], indicating that structural and functional
plasticity are tightly connected. In addition to structural plasticity, cAMP/cGMP pathways
were also reported to modulate functional synaptic plasticity and memory [59, 110, 111, 121,
123, 147, 151, 152, 173-175, 188, 196, 197]. Genetic and pharmacological manipulation of the
cAMP-PKA pathway consistently revealed that postsynaptic cAMP signaling mediates protein
synthesis-dependent late long-term potentiation (L-LTP) [59, 147] and long-term memory
consolidation [147, 175].
In contrast, the functional role of postsynaptic cGMP signaling at the hippocampus is not well
established. Mice lacking hippocampal cGMP-dependent PKG-I exhibit defective late-phase
potentiation, suggesting that cGMP also mediates L-LTP [111]. However, the approaches used
to study cGMP function in hippocampal synaptic plasticity did not specifically target
postsynaptic neurons, thus making it difficult to determine postsynaptic cGMP function
without its presynaptic contribution [121].
Reports of cGMP function in memory are consistent with cGMP enhancing synaptic
potentiation. Intra-hippocampal injection of a cGMP analog causes a dose-dependent
improvement in object recognition in rats, suggesting a role of cGMP in enhancing spatial
memory [197]. Likewise, the inhibition of cGMP-hydrolyzing enzymes improved object
recognition [151] and long-term memory formation in a social recognition task [152], thus
consistently implicating cGMP in improving memory. However, these experiments provide
61
limited information about the exact neuronal pathways and synapses involved in cGMP
signaling function.
The medial perforant path synapses of hippocampal DG neurons (PP-DG) exhibit post-synaptic
NMDA receptor-dependent long-term potentiation (LTP). In PP-DG synapses, postsynaptic
cAMP mediates protein synthesis-dependent late phase of potentiation (L-LTP) [99, 198], but
the postsynaptic function of cGMP is not known. Specific targeting of optogenetic guanylyl
cyclase (BlgC) to the DG (post-synaptic component of the PP-DG pathway) allows for light-
dependent synthesis of cGMP signal that is exclusively postsynaptic. This optogenetic
approach can be used to address the postsynaptic function of cGMP in PP-DG synaptic
plasticity. Furthermore, neurons in the DG are essential for learning and memory [199-204],
and are commonly utilized for optogenetic targeting [103].
To directly address the role of cGMP in synaptic plasticity and memory, I utilized the
optogenetic BlgC approach targeting PP-DG synapses. I generated cGMP by light and
investigated its effects on synaptic plasticity (in collaboration with the labs of Dr. John Roder
and Dr. Graham Collingridge at the LTRI) and mouse behavior (in collaboration with Dr. Keizo
Takao and Dr. Kaoru Inokuchi at the University of Toyama, Japan).
62
2.1 The role of postsynaptic cGMP in functional plasticity
cGMP enhances hippocampal potentiation through presynaptic mechanisms [121, 205]. Given
the postsynaptic expression of cGMP pathway components [68], it is unlikely that cGMP
effects on hippocampal LTP are exclusively presynaptic. Evidence from cultured hippocampal
neurons demonstrated that cGMP-dependent PKG-II regulates postsynaptic AMPAR
trafficking in CA3/CA1 synapses, suggesting a postsynaptic role of cGMP in synaptic
plasticity [123]. Hippocampal PP-DG NMDAR-dependent LTP is enhanced by postsynaptic
cAMP signaling [147], but the postsynaptic function of cGMP at PP-DG synapses is not well
understood.
To specifically target cGMP signaling to the postsynaptic cells of the PP-DG pathway, I virally
expressed the guanylyl cyclase enzyme (BlgC) in hippocampal DG granule neurons of the
mouse brain. To address the postsynaptic function of cGMP in hippocampal synaptic plasticity,
I prepared acute hippocampal slices from mice expressing BlgC in DG neurons and assayed
PP-DG synapse potentiation by recording field Excitatory Post-Synaptic Potentials (fEPSP)
with and without BlgC photoactivation. These experiments were performed in collaboration
with the labs of Dr. John Roder and Dr. Graham Collingridge at the LTRI.
Using these targeted optogenetic cGMP signaling approaches, I demonstrated a novel and
specific role of postsynaptic cGMP in enhancing synaptic potentiation at the dentate gyrus
synapses.
63
2.1.1 Validation of BlgC expression and light-dependent activation in DG neurons
To manipulate cGMP signaling by light, I first expressed the photoactivatable guanylyl cyclase
(BlgC-GFP) in DG granule neurons of 4-6 week old male C57B/6J mice. To deliver BlgC to
DG, I performed targeted stereotaxic viral particle microinjection of AAV (BlgC-GFP plasmid
under CaMKIIα promoter) into the living mouse brain (co-ordinates: -2.0mm AP, +/-1.3mm
ML, -1.9mm DV [206]) (Figure 24) and observed the expression of BlgC-GFP in acute
hippocampal slices by fluorescence imaging (Figure 24).
To validate the expressed BlgC-GFP photoactivity, I prepared and photoactivated hippocampal
lysates from mice expressing BlgC (5 minutes, blue light, 455 nm peak, 4.5 mW/mm2). Using
an ELISA assay, I demonstrated that activation of BlgC resulted in a robust light-dependent
cGMP increase, indicating light-dependent enzymatic functionality (Figure 25).
64
Figure 24: Schematic strategy of BlgC expression targeting the DG neurons of the
mouse brain
Schematic of the BlgC-GFP gene construct for viral delivery (left). Injection method for BlgC
targeting the dentate gyrus (center) and confocal images in acute hippocampal slices of mice
expressing BlgC-GFP-AAV1 (right). Scale bar is 100 µm.
65
Figure 25: Detection of cGMP produced by BlgC photo-activation in hippocampal slices
Hippocampal slice illumination for in vitro cGMP production (left) and ELISA cGMP
detection assay (center) were utilized to show light-dependent cGMP production in
hippocampal slices expressing BlgC with (+Light) and without light stimulation (-Light)
(right). The hippocampal slice lysate was illuminated by a 455 nm LED (4.5 mW/mm2, 5
minutes) (n = 3), p < 0.05, Unpaired t-test. All data are mean ± SEM.
66
2.1.2 Effect of postsynaptic cGMP on PP-DG L-LTP
To examine the postsynaptic function of cGMP at DG synapses, I prepared acute hippocampal
slices from mice expressing BlgC-GFP-AAV. Prior to each experiment I validated the
expression of BlgC by fluorescence microscopy. As a control, I utilized slices from wild-type
C57B/6J mice, which were also exposed to light to account for any effects of illumination.
To test whether the expression of BlgC in the DG affects basal synaptic transmission, I
recorded baseline fEPSPs in acute hippocampal slices from wild-type and BlgC-expressing
mice. I found that input/output (I/O) curves and DG-specific short-term paired-pulse
depression (PPR) at PP-DG synapses were comparable between wild-type and BlgC mice
(Figure 26), indicating that expression of BlgC at the synapse does not affect its function
without photoactivation.
To test whether exposure to light affects basal synaptic responses in PP-DG synapses of wild-
type and BlgC-expressing mice, I illuminated the slices with blue light (480 ± 15 nm, 1.5 mW,
4 minutes) and recorded fEPSPs before and after photoactivation (Figure 26). Light exposure
alone did not affect basal synaptic transmission at PP-DG synapses of wild-type or BlgC-
expressing mice, indicating that there is no immediate effect of cGMP on synaptic function.
These results indicate that the basal properties of PP-DG synapses are not affected by the
expression of BlgC, exposure to light, or exposure to cGMP (without potentiation).
67
Figure 26: Basal synaptic properties of wild-type and BlgC expressing mice
(A) Input/output relationship within wild-type dentate synapses (Black, n = 11 slices/10 mice)
and BlgC expressing synapses (Magenta, n = 14 slices/12 mice). (B) Comparison of paired-
pulse ratio in the PP-DG between wild-type and BlgC expressing mice. Paired-pulse activity
was calculated from the ratio of the second fEPSP slope to the first at different interpulse
intervals (wild-type n = 6 slices/6 mice, BlgC n = 6 slices/6 mice). (C-D) Quantification of the
baseline synaptic response with and without photoactivation. Slices were photoactivated (480
± 15 nm, 1.5 mW, 4 minutes) during baseline recording under a fluorescence microscope (BlgC
n = 6 slices/6 mice, wild-type n = 6 slices/6 mice). NS, p > 0.05, unpaired t-test. All data are
mean ± SEM.
68
To examine the direct effect of cGMP on L-LTP, I used hippocampal slices with and without
BlgC expression and induced PP-DG L-LTP by strong electric stimulation (tetanus: 4 x 100Hz)
in the presence of 10 µM bicuculline (to block GABAA-mediated inhibition [106]) either with
(+Light) or without (-Light) light (Figure 27).
Prior to each experiment, I validated the expression of BlgC in the DG by fluorescence
microscopy. Wild-type mice were used as a control for BlgC-expressing mice and the
photoactivation (light off/light on) experiments were paired and performed using hippocampal
slices from the same mouse.
I successfully induced L-LTP in hippocampal slices without expression of BlgC, which was
unaffected by exposure to blue light, indicating that photoactivation light did not affect L-LTP
(Figure 27). L-LTP in slices expressing BlgC without photoactivation was comparable to L-
LTP in slices without BlgC expression (Figure 27). In contrast, when BlgC was photoactivated
to increase post-synaptic cGMP at PP-DG synapses, L-LTP was significantly enhanced
(Figure 27), suggesting that cGMP may play a role in enhancing L-LTP.
69
Figure 27. Photoactivation of BlgC enhances functional long-term potentiation (L-LTP)
at PP-DG synapses
(A) Time course of long term potentiation (L-LTP), induced by a strong tetanic stimulation (4
x 100Hz, 3x pulse width, 10 second interval, + 10 µM bicuculline) in wild-type mice with
(+Light, wild-type; grey; n = 6 slices/6 mice) and without (-Light, wild-type; black; n = 6
slices/6 mice) illumination by blue light (480 ± 15 nm, 5 min, blue bar). L-LTP is compared as
average of 80-90 min after tetanic stimulation (bar graph, right) and shows no significant effect
of light on L-LTP in wild-type mice. n=6 slices/6 mice. NS, p > 0.05, paired t-test. All data are
mean ± SEM. (B) L-LTP in hippocampal slices expressing BlgC with (+Light, wild-type; grey;
n = 6 slices/6 mice) and without (-Light, wild-type; black; n = 6 slices/6 mice) illumination by
blue light (480 ± 15 nm, 5 min, blue bar). L-LTP measured as average of 80-90 min after tetanic
stimulation (bar graph, right) shows significant enhancement of L-LTP in the light-stimulated
condition p = 3.35E-09, paired t-test. All data are mean ± SEM.
70
To confirm that the cGMP effect is postsynaptic, I measured the paired pulse ratio (PPR) in
hippocampal slices with and without BlgC expression during L-LTP either with (+Light) or
without (-Light) light (Figure 28). I found no significant difference in PPR before and after L-
LTP, indicating that the cGMP effect on L-LTP is postsynaptic. Taken together, these results
indicate that postsynaptic cGMP significantly enhances functional potentiation in the PP-DG
pathway.
71
Figure 28: The effect of BlgC photoactivation on paired pulse ratio before and after
tetanic stimulation
Comparison of paired-pulse ratio before (-10 to 0 min, black) and after tetanic stimulation (80
to 90 min, white) ± blue light in both wild-type and BlgC expressing hippocampal slices (B).
NS, p > 0.05, paired t-test. All data are mean ± SEM.
72
2.2 The role of cGMP in mouse behavior
Hippocampal memory can be modeled in mice using a variety of behavior tests thought
to represent the human condition. Episodic memory (what/where/when) [207] is further
subdivided into short-term “working memory” of recently learned information and training-
dependent “reference memory” of information from previous trials [208]. Maze tests that rely
on training trials can be used to evaluate reference memory (eg. Barnes Maze) [137]. Other
maze tests (such as T-maze Forced Alternation) and Object location tests can be used to assay
short-term information processing, or working memory [138, 139]. Mice lacking cGMP-
dependent protein kinase G-I (PKG-I) exhibit impaired reference memory [209] and those
lacking cGMP-dependent PKG-II show impaired working memory and increased anxiety [210-
212]. Increasing cGMP signaling by blocking its degradation causes improved reference
memory and increased anxiety [213-215]. This global manipulation of cGMP suggests that
cGMP improves memory through PKG and has variable effects on anxiety.
However, the function of cGMP in the dentate gyrus, which is involved in memory [199],
Alzheimer’s disease [200, 201], anxiety [202, 203] and social behavior in mice [204], is not
known. BlgC optogenetics can be targeted with excellent spatial and temporal precision to
manipulate cGMP by light at various time points and brain segments. Therefore, to directly
address the function of cGMP in mouse behavior, I utilized BlgC to manipulate cGMP in
hippocampal DG granule cell neurons and performed a battery of mouse behavioral tests,
including tests of anxiety, sociability and memory. I found that cGMP is associated with an
enhancement in anxiety and reference memory, and a slight enhancement in sociability. These
experiments were performed in collaboration with professor Takao at Toyama University
(behavior test training) and professor Inokuchi (surgery training).
73
2.2.1 BlgC expression and light-dependent activation in the mouse brain
To express and photoactivate BlgC in the DG of living mice, I microinjected BlgC-GFP-AAV
targeting the DG of 26 C57/Bl6 wild type mice (bilateral) and surgically equipped the same
mice with either a bilateral fiberoptic LED system (n=14) or, as a control, a “dummy” optic
fiber, which produces no light but exhibits the same physical properties (n=12) (Figure 29).
All mice were housed in a Home Cage Monitoring System with automated video tracking and
habituated to wearing a detachable battery (necessary to power the LED).
74
Figure 29: Schematic of light-dependent cGMP signal delivery in the mouse brain
AAV1 construct carrying BlgC-GFP is injected into the mouse brain targeting the dentate gyrus
(top), then an LED (middle right) or “dummy” (middle left) optic fiber is surgically implanted
for cGMP and control mice, respectively. Bottom: hippocampal slice showing BlgC-GFP and
LED light targeting neurons of the dentate gyrus for localized cGMP signaling.
75
2.2.2 Effect of postsynaptic cGMP on mouse behavior
In collaboration with Dr. Keizo Takao at the University of Toyama and his experienced
technician Adachi-san, I designed and carried out a battery of behavioral tests including general
health and neurological screening, anxiety, social interaction and memory tests (Figure 30).
For each behavior test, we carefully selected the light-activation period to examine the effects
of cGMP without interfering with the integrity of the test. Some tests, such as the pre-pulse
inhibition (PPI) test, were excluded because the holding chamber is not able to accommodate
a mouse equipped with an LED implant and battery. After the behavior testing was complete,
I sacrificed all mice and observed the expression of BlgC, which resulted in the exclusion of 1
BlgC-cGMP and 2 dummy mice from the results pool.
76
Figure 30: Summary schematic of mouse behavior analysis
Schematic and time course of the mouse behavior tests performed using BlgC-expressing mice
surgically equipped with LED and dummy implants for photo-activation of BlgC.
77
2.2.2.1 General Health and Neurological Screening (GHNS)
After a 2-week recovery period following surgery, I carried out general health and neurological
screening, rotarod and hotplate tests to determine the health condition of the test mice and
confirm their ability to perform the remaining tests. The general health and neurological
screening indicated no seizures or abnormalities in righting reflex, body weight, temperature
or grip strength, therefore confirming the mice are fit to participate in behavioral experiments.
In the rotatrod and hot plate tests, the latency to fall and response to temperature increase were
indicative of good motor function and sensitivity in these mice (Figure 31).
To examine the effect of cGMP on mouse condition and performance, I included a 5-minute
BlgC photo-activation prior to the screen, mimicking the same photoactivation protocol I used
in electrophysiology. cGMP signaling had no significant effect on general health and
neurological conditions, indicating that cGMP has no effect on basal performance in these
mice.
78
Figure 31: General Health and Neurological Screening (GHNS)
Measurements of body weight (A), body temperature (B) and strength using grip (C) and wire
hang (D) methods show no significant abnormalities in mice 2 weeks after surgery. Rotarod
test (E) and hotplate test (F) show no abnormalities in motor function and sensitivity of mice.
Cntrl n=10, LED n=13, data are mean ± SEM, Statistical analysis was performed using an
ANOVA.
79
2.2.2.2 Anxiety Tests (LD, EP, OF)
Next, I tested anxiety by measuring exploration time in undesirable conditions, such as bright
light (Light-dark test, LD, Figure 32), elevated open areas (Elevated plus maze, EP, Figure
33) and open bright areas (Open field test, OF, center time, Figure 34). The control and cGMP
mouse cohorts were counterbalanced such that each test was performed on an equal number of
control and cGMP mice simultaneously (in the case of multiple chambers). In single chamber
tests, such as the Elevated plus maze, a control mouse test was followed by cGMP mouse test
until all mice were evaluated. cGMP was associated with an increase in anxiety, which was
especially pronounced in the last 30 minutes of the test (Figure 34). The delayed effect of
cGMP in the late stage of this 120-minute may explain the lack of effect in other anxiety tests
that last a total of 10 minutes.
80
Figure 32: Light-Dark (LD) test
Photo (A) and light-activation timeline (B) during LD test. Control and cGMP mice show a
preference for the dark compartment (C), but no significant differences in distance travelled
(C), stay time in light (D), number of transitions between compartments (E) or latency to light
(F) between the two groups. Cntrl n = 10, cGMP n = 13, data are mean ± SEM, Statistical
analysis was performed using an ANOVA.
81
Figure 33: Elevated Plus Maze (EP) test
Photo (A) and light-activation timeline (B) during EP test. Control and cGMP mice show no
significant difference in the number of entries into open arms (C), percent of entries into open
arms (D), total distance travelled (E) or the percent of time spent in open arms (F) of the
elevated maze. Cntrl n = 10, cGMP n = 13, data are mean ± SEM, Statistical analysis was
performed using an ANOVA.
82
Figure 34: Open field (OF) test
Photo (A) and light-activation timeline (B) during OF test. Control and cGMP mice show no
significant difference in the total distance travelled (C). During the last 30 minutes of the test,
the cGMP mice spend significantly less time in the center of the arena, compared to control
mice (D, p = 0.0308). Overall (0 min-120 min), there is no significant difference in center time
(D, p = 0.0589). Cntrl n = 10, cGMP n = 13, data are mean ± SEM, Statistical analysis was
performed using an ANOVA.
83
2.2.2.3 Sociability Tests (CSI, SI)
I next examined the effect of cGMP on sociability using Crawley’s social interaction test (CSI),
where mice are presented with a caged stranger vs an empty cage (Figure 35). 2 control and 2
cGMP mice were tested simultaneously in this test, and there was no significant effect of cGMP
on the time that mice spent interacting in this test.
In the single chamber social interaction test (SI), subjects freely interact with an uncaged mouse
from the same cohort. Control and cGMP mice were weight-matched prior to testing to avoid
aggressive behavior and were equipped with dummy and LED battery (respectively) for the
duration of the test. cGMP signaling was associated with a 39% increase in the number of
contacts and a 33% increase in contact duration, suggesting that cGMP signaling in the DG
may be associated with increased sociability (Figure 36).
84
Figure 35: Crawley’s social interaction (CSI) test
Photo, with schematic of the two trials (A) and light-activation timeline (B) during CSI test.
Control and cGMP mice show a significant preference for the cage with a stranger mouse
compared to empty cage (C). The social preference index was calculated for each group (Time
exploring stranger cage/(time exploring empty cage + Time exploring stranger cage)) and
showed no significant difference between cGMP and control mice (D). Similarly, both groups
showed preference for the novel stranger mouse (E), but no significant difference in Social
novelty preference index between the two groups (F). Cntrl n = 10, cGMP n = 13, data are
mean ± SEM, Statistical analysis was performed using a paired t-test.
85
*
Figure 36: Single chamber social interaction (SI) test
Photo (A) and light-activation timeline (B) during SI test. cGMP mice show a significant
increase in the total duration of contacts (C), and total number of contacts (D) with a stranger
mouse, compared to controls. There were no significant differences between the two groups in
the total duration of active contacts (E) and total distance travelled (F). Cntrl n = 10, cGMP n
= 13, data are mean ± SEM, Statistical analysis was performed using an ANOVA.
86
2.2.2.4 Memory Tests (OLT, BM, TM)
I tested spatial memory using Object Location Test (OLT), Barnes Maze (BM) and T-Maze
(TM). The control and cGMP mouse cohorts were counterbalanced such that each test was
performed on an equal number of control and cGMP mice simultaneously. In OLT, both
cohorts of mice (control and cGMP) spent more time around the novel object, indicating
memory formation. However, there was no significant difference in the discrimination index
between the two cohorts, suggesting that cGMP had no significant effect on spatial/working
memory during OLT (Figure 37).
To test the role of cGMP in training-dependent reference memory, I next incorporated cGMP
photo-activation into the daily training trials of the Barnes Maze (BM) memory test (Figure
38). The mice were trained in the same order each day (counterbalanced- control mouse, then
cGMP mouse), then returned to their home cage. After 1 training trial was complete for all
mice, then the training was repeated (in the established order) for all mice for up to 3 training
trials/day. No trial was started if it could not be completed for all mice that day. After 12 trials,
I found that cGMP mice made fewer errors, spent less time and travelled a shorter distance to
find the escape hole compared to controls (Figure 38), suggesting that cGMP signaling
improves reference memory. In probe test 1 (PT1) (Figure 39), which was instituted 24 hours
after the last training, cGMP had no significant effect on long-term memory.
In T-Maze Forced Alternation (TMFA), the mice were offered a food reward in alternating
arms of the maze, depending on the trial [138] (Figure 40). The order of mice being evaluated
was established in advance (randomized and counterbalanced) and maintained each day. No
trial was started if it could not be completed for all mice that day. There was no significant
87
difference in the total distance travelled, the latency, or the correct % of responses between
cGMP and control mice, suggesting that cGMP does not affect spatial working memory in this
task. When a delay was introduced between each arm alternation (3s-60s), there was no
significant difference in the percentage of correct responses between control and cGMP mice.
88
Figure 37: Object Location test (OLT)
Photo (A) and light-activation timeline (B) during OLT test. During familiarization phase, mice
are introduced to the original configuration of two identical objects, and cGMP signaling is
activated by light in cGMP mice. Data shown here is the first 5-minute averages of test phase,
where one object is displaced (object in novel location, black bars). Both cGMP and control
mice show a preference for novelty based on time spent around displaced object (C,D) and
number of entries to the novel object region (E,F). The discrimination index (G) (time around
novel object/(time around novel object + time around familiar object)) shows no significant
difference between control and cGMP groups. Cntrl n = 10, cGMP n = 13, data are mean ±
SEM, Statistical analysis was performed using an ANOVA and a paired t-test for
discrimination index.
89
Figure 38: Barnes Maze (BM) training trials
Photo (A) and light-activation timeline (B) during BM test. During trials 1-15 mice are trained
to learn the location of the escape box; learning is measured as latency(s), distance travelled
(cm) and error to target hole. (C) cGMP mice show a significantly lower latency (D) distance
travelled (E) and error when finding the target hole during the last 3 training trials. Insets are
average of trials 12-15. Cntrl n = 10, cGMP n = 13, data are mean ± SEM, Statistical analysis
was performed using an ANOVA.
90
Figure 39: Barnes Maze (BM) probe test (PT1)
24h after the last training trial, mice were subjected for a probe test (PT1) and allowed to
explore the arena for 3 minutes with the escape box removed. Total time spent around each
hole was recorded (sec). There is no significant difference in the time spent around the target
hole between cGMP and control mice. Cntrl n=10, cGMP n=13, data are mean ± SEM,
Statistical analysis was performed using an ANOVA.
91
Figure 40: Forced Alternation T-maze test (TMFA).
Photo (A) and light activation timeline (B) during Forced Alternation T-maze test. After a 1-2
week food restriction phase mice reach ~80-85% of their original body weight and are
considered motivated enough for the food reward (sucrose pellets). Pellet is delivered to
alternating arms while the correct responses, distance and latency of each mouse is recorded.
During delay trials, a 3s, 10s, 30s or 60s delay is inserted between each arm alteration task.
There is no significant difference in the correct responses (%) (A), latency (s) (B) or distance
travelled (C) between mice stimulated to produce cGMP (Bl, Black bars) and control mice
(Dm, white bars) during trials 1-5. When a delay is introduced between each arm alternation
(3s-60s, D-F), there was no significant difference in the percentage of correct responses
between control and cGMP mice. Cntrl n = 10, cGMP n =13, data are mean ± SEM, Statistical
analysis was performed using an ANOVA.
92
Altogether, cGMP signaling is associated with increased sociability, improved reference
memory and increased anxiety (more pronounced in the last 30 minutes of the test) (Table 1).
The effects of cGMP on anxiety and memory are not evident in shorter tests (10-minute
protocols), but only in the open field test (120-minute duration).
93
Table1: Summary of the mouse behavior testing battery
Summary of the mouse behavior testing battery showing the effect of light-activated cGMP
production on memory, anxiety and sociability
94
2.3 Discussion
Using targeted optogenetic cGMP signaling approaches, I examined the function of
cGMP at the level of the synapse, neuron and the brain. To determine a functional role of
cGMP, I targeted the PP-DG synapses and found that postsynaptic cGMP enhances synaptic
potentiation and reference memory in mice.
Postsynaptic cGMP function can be difficult to target in synaptic plasticity studies [113].
Therefore, a key advantage of using optogenetic cGMP signaling is the capacity to target
specifically postsynaptic pathways. Precisely localized delivery of BlgC-GFP-AAV1 to the
DG ensures that cGMP signal is produced by light only in the DG granule cells, which are the
postsynaptic component of the PP-DG pathway. Therefore, I carefully optimized the virus
delivery protocol for BlgC-GFP expression in the DG and validated its photoactivation in
hippocampal lysates using an ELISA. Another key benefit of optogenetics is the opportunity
to choose the time points and duration of cGMP signaling. I photoactivated BlgC to produce
cGMP for 5 minutes (4 minutes before potentiation and 1 minute during), in order to maximize
its effect on plasticity and determine whether cGMP affects initial potentiation, late plasticity
or both. Without altering the baseline synaptic properties, light-dependent cGMP signaling
significantly enhanced synaptic plasticity, increasing both early and late functional synaptic
potentiation. These findings indicate the involvement of post-synaptic cGMP in hippocampal
synaptic function, and suggest that cGMP enhances synaptic strength.
The finding that cGMP signaling generates a long lasting enhancement of synaptic plasticity
(>90 minutes), suggests that cGMP may participate in protein synthesis-dependent pathways
at PP-DG synapses. Previous reports indicate that cGMP-dependent PKG may act in parallel
with PKA to increase phosphorylation of the transcription factor CREB and promote CREB-
95
mediated gene expression in the CA1 region [216] and in neurons of the lateral amygdala [217].
The robust light-dependent effect of cGMP on synaptic plasticity also validated the use of BlgC
enzyme to efficiently increase cGMP signaling in the mouse brain.
I next utilized BlgC optogenetics to study the effect of targeted cGMP signaling on mouse
behavior. These experiments were performed in collaboration with Dr. Keizo Takao’s
laboratory in Japan and funded by the Weston Brain Institute International scholarship. With
the help of Dr. Takao, I carefully orchestrated the order and BlgC photoactivation timeline of
each test to minimize stress to the animal and maximize the effect of cGMP. Whenever
possible, I photoactivated BlgC for 5 minutes to mimic the electrophysiology experiments.
During anxiety tests with a 10-minute duration (Light-dark test and Elevated Plus maze)
control and cGMP mice performed equally. However, in the Open Field test (2-hour duration),
cGMP mice exhibited signs of increased anxiety, especially in the last 30 minutes. These results
suggest that effects of cGMP on anxiety are delayed, which is consistent with the prediction
that cGMP effects in the DG are protein synthesis-dependent. Interestingly, even though the
single chamber Social Interaction test lasts only 10 minutes, cGMP signaling was associated
with increased sociability. This finding suggests that cGMP may carry out more than one
function in the DG, affecting neurons involved in anxiety in the long-term and sociability in
the short-term.
To examine the role of cGMP in memory at the DG, I performed the (1) “Object Localization
test”, which evaluates novelty and spatial working memory, (2) training trial-dependent
“Barnes maze test” of long term reference memory and (3) “T-Maze Forced Alternation test”
of working memory. I saw no significant effect of cGMP on short-term working memory in
either Object Localization test, or T-Maze test. Furthermore, there were no significant
96
differences in maze performance between control and cGMP mice in the first 12 trials of the
Barnes Maze. However, in trials 12-15, cGMP mice showed significantly improved reference
memory, compared to control. This delayed improvement in reference memory is consistent
with cGMP function in enhancing synaptic plasticity. 24 hours after the last training trial, mice
were tested in a probe test of long-term memory, and even though cGMP mice performed better
in trials 12-15, there was no significant difference in the probe test (PT1). The motivating factor
in the Barnes maze test is anxiety, as the mice are searching for a dark escape hole to hide from
the open arena. Therefore, I cannot rule out the possibility that cGMP mice are finding the
escape hole faster because of their enhanced anxiety and not because of improved memory. In
the future, other trial-dependent tests of reference memory, such as the T-maze Left/Right
discrimination test, can be used to verify the role of cGMP.
In summary, I utilized BlgC to generate cGMP by optogenetics in living brain tissue and found
that cGMP signaling in the DG enhances synaptic function as well as anxiety, sociability and
reference memory. Upregulating cGMP signaling was reported to enhance memory in
neurodegenerative disorders such as Alzheimer’s disease [153-155], suggesting that cGMP
signaling is impaired. Therefore, optogenetic cGMP manipulation can be used in combination
with mouse models of Alzheimer’s disease to discover new therapeutic targets and treatment
approaches for this disease.
97
Conclusion and Future Directions
cGMP signaling plays a role in activity-dependent structural and functional modulation
of the synapse, which is essential for learning and memory. However, little is known about the
postsynaptic function of cGMP in structural potentiation and plasticity at the dentate gyrus. To
address this, I characterized a novel optogenetic method for targeted cGMP manipulation and
revealed the effect of postsynaptic cGMP signaling at the level of the synapse, neuron and the
brain.
In Chapter 1, I established a two-photon optogenetic approach to non-invasively manipulate
cGMP signaling at the level of the synapse and demonstrated that postsynaptic cGMP and
cAMP competitively govern structural potentiation of dendritic spines. cGMP does not affect
basal structural potentiation, but negatively regulates cAMP-dependent structural enhancement,
suggesting a bidirectional postsynaptic regulation mechanism that is supplementary to basal
sLTP and mediated by competitive cAMP/cGMP signaling. I also demonstrated that PKG is
critical for negative regulation of the cAMP-enhanced synaptic structure, indicating the
involvement of the cGMP-PKG pathway. Future experiments will further elucidate this cAMP-
cGMP interplay and identify the cross-over mechanism.
In the future, it would also be valuable to examine the relationship between cAMP/cGMP
concentration and effects on the dendritic spine structure. Since forskolin and BlgC produce
variable concentrations of cAMP and cGMP, it may be difficult to conclude the true nucleotide
function at physiological concentrations. Therefore, it will be useful to correlate
photoactivation parameters of BlgC with the concentration of cGMP generated during the
experiment. Furthermore, instead of forskolin, it is possible to produce cAMP by PAC
(Photoactivatable Adenylyl Cyclase), an enzyme characterized in our lab for optogenetic
98
cAMP signaling at the dendritic spine. The combination of these methods can further enhance
our understanding of the novel interplay between cAMP and cGMP during structural
potentiation, and help us reveal the physiological function of the crosstalk between them.
In Chapter 2, I adapted the optogenetic method to manipulate postsynaptic cGMP levels in
hippocampal slices and provided the first report of postsynaptic cGMP function in the dentate
gyrus LTP. I revealed that cGMP signaling immediately prior to synaptic potentiation
significantly enhances PP-DG L-LTP, suggesting a function of cGMP in the late protein
synthesis-dependent plasticity pathways. In the future, this optogenetic photoactivation
approach can be adapted to further refine cGMP signaling function in plasticity by
manipulating cGMP concentration (changing duration and frequency of photoactivation) and
signaling time course (photoactivating before vs after potentiation is induced). To further
elucidate the synaptic plasticity signaling pathways, it is also possible to demonstrate the main
cGMP effectors using specific cGMP pathway inhibitors (such as PKG). However, this
pharmacological approach presents other potential complications, since inhibitors of cGMP-
dependent protein kinase (PKG) were reported to block the induction of LTP [115, 218]. Our
lab has also generated light-activated phosphodiesterase enzymes that lower cGMP levels by
light-induced degradation. These enzymes can be used in the future to determine how
decreasing endogenous post-synaptic cGMP affects synaptic plasticity.
I also used optogenetics to target the DG region in living mice and found that localized cGMP
signaling is associated with a delayed enhancement in anxiety and reference memory,
suggesting a role of cGMP signaling in long-term behavioral expression. cGMP signaling had
no effect in short-term anxiety tests. In the Open field test, which has a 120-minute duration,
cGMP increased anxiety, especially in the last 30 minutes of the test. In the future, it is possible
99
to confirm the effect of cGMP on anxiety by photoactivating cGMP signaling 120 minutes
prior to the start of the shorter anxiety tests (Light Dark test and Elevated Plus maze).
The excellent spatial and temporal control of BlgC photoactivation introduces new possibilities
for studying memory in the future. For example, targeting BlgC optogenetics to the amygdala
will make it possible to address the function of cGMP in established paradigms such as fear
memory [219]. The opportunity to control the time window when cGMP is produced can also
be adapted to test the function of cGMP at various stages of memory processing. For example,
in an object localization test, cGMP signaling can be exclusively photoactivated during the
familiarization phase (to test the effect on memory acquisition), during the interval period (to
test the effect on consolidation) or during the test phase (to study cGMP effects on memory
retrieval).
cGMP metabolizing enzymes are also commonly targeted for the treatment of Alzheimer’s
disease [153-155], suggesting that BlgC can be used in combination with Alzheimer’s disease
mouse models to investigate novel therapeutic targets and treatment approaches in the future.
My project was the first to apply optogenetic approaches to address post-synaptic cGMP
function in structural synaptic plasticity and dentate gyrus synaptic plasticity. The cGMP
optogenetic methods established here can provide a novel approach to elucidate the dynamic
molecular mechanisms of synaptic plasticity and memory in health and disease.
100
Materials and Methods
3.1 DNA constructs GFP (EGFP with A206K mutation to prevent dimerization [220]) was subcloned into the
pCAG plasmid vector[221]. Photoactivatable guanylyl cyclase (BlgC), was generated by
altering photoactivatable adenylyl cyclase PAC (optimized for human expression, addgene ID
28134) [222]. Three amino acids in the ATP binding site of PAC were altered for GTP binding
(K197E, D265K, T267G) [177], which was fused with RFP (tdTomato) [192] as a red
fluorescent marker at the N-terminal and subcloned into a pCAG plasmid vector [221]. The
cGMP FLIM probe (cGiR) was constructed from the Gln79 – Tyr345 in cGMP-dependent
protein kinase I (cGKI) [193] of Cygnet 2.1 (addgene ID 19737) [223] and subcloned into a
pCAG plasmid vector after replacing CFP with RFP (tdTomato).
3.2 Animal care Organotypic slice cultures of rat hippocampus and acute slices of mouse hippocampus were
prepared in accordance with the guidelines of the University Health Network (UHN) Animal
Care Committee (Toronto, Canada). Stereotaxic microinjection surgery protocols were
approved by the Toronto Centre for Phenogenomics (TCP) Animal Care Committee (Toronto
Canada) and the University of Toyama, (Toyama, Japan). Animal procedures for mouse
behavioral assays were approved by the University of Toyama, Japan.
3.3 Two-photon imaging and uncaging In organotypic hippocampal slices, CA1 pyramidal neurons were transfected biolistically at
DIV 5 (GFP volume filler, or other constructs as specified), and live imaging experiments were
performed 3-5 days after transfection in the distal regions of the main apical dendrite (860 nm
101
two-photon excitation, 495-540 nm for GFP imaging with Olympus FV1000MPE). The
cultured slices were continuously perfused with ACSF (artificial cerebrospinal fluid) in the
microscope chamber at 30 °C equilibrated with 5% CO2/95% O2 [22]. In all experiments, spine
size over time was measured by summed GFP fluorescence intensity of the spine volume from
15-20 sections taken at 0.5 µm intervals of z-stack images of the dendrites using Imaris
software (Bitplane, Zurich, Switzerland).
3.3.1 Structural potentiation
ACSF for LTP induction by HFS (10 mW, 4 msec, 1 Hz, 60 times) contained 119 mM NaCl,
2.5 mM KCl, 4 mM CaCl2, 26.2 mM NaHCO3, 1 mM NaH2PO4 and 11 mM glucose, 1 µM
tetrodotoxin, and 2.5 mM 4-methoxy-7-nitroindolinyl (MNI)-L-glutamate (Tocris, Bristol,
UK). The two-photon glutamate uncaging (MNI-glutamate, Tocris, Bristol, UK) at 720 nm
was applied near the targeted spine (~1 µm from the tip of the spine head), and subsequent
images of dendritic spine enlargement were captured using two-photon GFP imaging (900 nm
excitation, 495-540 nm emission). Results for each experiment are a collection of data from
several dendritic spines from multiple neurons targeted on one or more hippocampal slices,
depending on expression.
For uncaging in the presence of cAMP, ACSF solution was supplemented with 50µM of
Forskolin (Calbiochem, San Diego, CA) and perfused for 5 minutes total (4 minutes
immediately prior to uncaging and during 1 minute of uncaging). After 5 minutes, slices were
perfused again in ACSF without forskolin in order to avoid damage from excessive forskolin
exposure.
102
3.3.2 Depotentiation of cAMP-dependent structural enlargement
For depotentiation of the enhanced synaptic structure, the same steps were carried out as above
to induce cAMP - LTP, but immediately after HFS, the same spine was subjected to LFS
(0.1Hz, paired pulses with 0.5ms duration of activation and 50ms intervals, 15mW power for
15 minutes) while being perfused with new ACSF containing 119 mM NaCl, 2.5 mM KCl, 2
mM CaCl2, 26.2 mM NaHCO3, 1 mM NaH2PO4 and 11 mM glucose and 2.5 mM 4-methoxy-
7-nitroindolinyl (MNI)-L-glutamate.
3.3.3 Two-photon optogenetic cGMP production
For optogenetic cGMP signaling at the target dendritic spine, RFP-BlgC and GFP (volume
filler) were cotransfected at a ratio of 4:1. To photoactivate postsynaptic BlgC, the dendritic
spines were excited at 1,000 nm two-photon laser light (point-scanning, 11mW, 30 sec)
immediately prior to HFS. In all experiments, spine size over time was measured by summed
GFP fluorescence intensity of the spines from 15-20 sections taken at 0.5 µm intervals of z-
stack images of the dendrites using Imaris software (Bitplane, Zurich, Switzerland).
103
3.4 BlgC photoactivation 3.4.1 cGMP detection by ELISA in vitro
For expression of the photoactivatable enzyme BlgC, 5 x 105 HEK293 cells were seeded into
each well of 6-well plates and transfected with BlgC (2.5 µg plasmid vector) using
Lipofectamine 2000 (Life Technologies, Grand Island, NY). The cells were harvested 48 hours
after transfection, homogenized in a buffer (40 mM HEPES/Na, pH 8.0, 0.1 mM EGTA, 5 mM
magnesium acetate, 1mM DTT, and 0.01% Tween-20) by sonication, and centrifuged at 16,000
g for 15 min to clear large cell debris. The supernatant was isolated and stored -80 freezer for
experiments. For the in vitro photoactivation assay, the cell lysate were excited with a 455 nm
LED (4.5 mW/mm2; ThorLabs, NJ, USA) on a plastic paraffin film covered glass slide at room
temperature, and the synthesized cGMP levels were measured by a cGMP ELISA kit (Enzo
Life Sciences, NY, USA) [177, 222]. During photoactivation, cell lysates were resuspended in
the reaction solution (100 µl) containing 40 mM HEPES/Na, pH 8.0, 0.1 mM EGTA, 5 mM
magnesium acetate, 1mM DTT, 0.01% Tween-20 and 200 µM MgGTP. Excitation with two-
photon laser light (30 mW, 700 – 1025 nm) was performed under the 60X objective lens in a
custom-made two-photon microscope (FV1000 MPE; Olympus, Tokyo, Japan) equipped with
2 two-photon lasers (Mai Tai HP DeepSee; Newport, CA, USA). The light-dependent reactions
were performed in dark and stopped at the appropriate time points by application of 0.1 M HCl
following the ELISA kit protocol.
104
3.4.2 cGMP detection by FLIM
For in vitro cGMP detection by FLIM, HEK293 cell lysates expressing cGiR probe with and
without BlgC were placed in a microscope chamber to measure lifetime change in cGiR upon
photoactivation of BlgC by focused blue light (414 ± 21 nm, mercury arc lamp) under a
microscope.
To detect BlgC photoactivation in the neurons, cGMP FLIM probe was biolistically transfected
with or without BlgC in CA1 pyramidal neurons in organotypic cultured hippocampal slices
as described previously. The expressed BlgC was photoactivated by single photon light (414 ±
21 nm, mercury arc lamp, 60 sec) in the neurons or two-photon laser light (1000 nm, 30 sec)
at the target dendritic spines. The lifetime change of cGiR FLIM probe was measured by a
two-photon FRET/FLIM microscope equipped with a time-correlated single photon counting
(TCSPC) system (PicoHarp 300, FLIM upgrade kit for Olympus FV1000MPE, SymPhoTime
software; Picoquant, Berlin, Germany) using single plane of YFP fluorescence (520-542 nm)
images (donor of cGiR: 900 nm excitation, 30 sec repetitive FLIM imaging time). Fluorescence
lifetime data were analyzed by SymPhoTime software (PicoQuant, Berlin, Germany)
[224]. For average time-domain fluorescence lifetime measurements, the lifetime decay curves
were fit by a double exponential decay model using the tail-fitting analysis [224, 225]. The
fluorescence lifetime pseudo-colour image of neurons was constructed using a pixel by pixel
fitting function.
105
3.5 Pharmacology Drugs were prepared as stock aliquots of 1000x final concentration and stored at -20C. 2µM
PKG inhibitor KT 5823 [189, 226], 100nM PDE2 inhibitor Bay 60-7550 [150] and 100µM
CNG channel inhibitor L-cis-diltiazem [76] were diluted to final concentration in carbogenated
ACSF and bath applied immediately prior to the experiment.
3.6 AAV packaging High titer (1E+13) AAV particles (serotype 1) were produced by SignaGen Laboratories
(Rockville, MD, USA) containing CaMKIIα promoter-EGFP-BlgC construct.
3.7 Surgery: Stereotaxic microinjection and LED implant Mice were anesthetized with isofluorane (2% vol/vol) and then transferred to a stereotaxic
apparatus. Mice were secured to the apparatus by fixing ears bars to the head and insertion of
the incisor adaptor. A nose cone was used to administer isofluorane during surgery to maintain
anesthesia as confirmed by mild pinching of the tail. Mice were injected with 0.1 mL of
meloxicam (0.5 mg/mL) and 0.1 mL of saline subcutaneously prior to surgery. Hair atop the
scalp was removed with electric clippers and antiseptic solution was applied topically to clean
the skin. Additionally, Tear-Gel (Alcon, TX, USA) was applied to lubricate the eyes and
reapplied when necessary. A scalpel was used to make a midline incision and expose the
bregma and lambda landmarks of the skull. After leveling the head, a drill was used to make
two small holes 2 mm posterior from bregma and 1.3 or 1.6 mm lateral to the midline. Injection
volumes were 2 µL per hemisphere and delivered at a rate of 0.1-0.2 µL per minute using a
gastight syringe with a 26-gauge needle (Hamilton, Reno, NV, USA). Following injection, the
syringe was kept in place for an additional five minutes to allow diffusion of the virus particles.
The syringe was then slowly withdrawn from the head and the injection site was wiped clean.
106
The skin was sutured and antibiotic ointment was applied over the wound. The mouse was
removed from the stereotaxic apparatus and placed in a recovery cage under a heat lamp.
Meloxicam was administered subcutaneously every 24 hours for 2 to 3 days following surgery.
Mice were given at least 3 weeks of postoperative recovery.
For behavioral testing, mice were microinjected with virus and then surgically equipped with
a permanent bilateral LED (or dummy) implant targeting the same site where virus was
injected. A bilateral 470 nm wavelength LED device (0.75 mm diameter, 20 mW, Eicom, CA,
USA) was inserted slowly to minimize damage to brain tissue. The LED device was fixed to
the skull using two kinds of dental cement (Bistite II DC, Tokuyama Dental America Inc.,
Japan, and UNIFAST Trad Liquid, GC AMERICA INC, USA) and cured. The mouse was then
removed from the stereotaxic device and placed in a recovery cage under a heat lamp.
Analgesia (METACAM) was continued by subcutaneous injection every 24 hours for 3 days
post-surgery. At least 14 days were interspaced between surgery and behavior testing.
3.8 Fluorescence imaging of mouse hippocampal slices Acute mouse hippocampal slices of 400 µm thickness were prepared and bathed in ACSF
equilibrated with 95% O2 and 5% CO2 at room temperature. GFP-BlgC fluorescence imaging
was conducted using a Nikon C2+ confocal laser scanning microscope with 4X objective lens
(Nikon, Tokyo, Japan) and 488 nm laser excitation (514 nm emission). Image analysis and
processing was performed using NIS-Elements software (Nikon, Tokyo, Japan).
3.9 Electrophysiology Acute hippocampal slices were prepared as previously described [105] from BlgC-expressing
mice and their wild-type littermates. After 1.5-2 hours recovery, slices were transferred to a
107
chamber perfused with an ACSF solution containing 124 mM NaCl, 3 mM KCl, 2.5 mM CaCl2,
1.3 mM MgSO4, 1.25 mM NaH2PO4, 26 mM NaHCO3, 10 mM glucose (pH 7.4, 30°C, 1.5
ml/min) equilibrated with 5% CO2/95% O2. Recordings of field excitatory postsynaptic
potentials (fEPSPs) were conducted as previously described [105] in the DG (perforant
path/dentate synapses) while blocking inhibitory synaptic function using 10 µM Bicuculline
[106]). The stimulation electrode was positioned in the dorsal blade of the dentate molecular
layer for medial perforant path stimulation. Paired field responses were evoked by stimulating
with an intensity (0.05 msec pulses, 40 msec apart) that yielded fEPSPs that were 40 % of the
maximum spike-free fEPSP size. Responses were evoked and acquired every 20 sec throughout
the experiment using an Axopatch 1D amplifier (Axon Instrument) digitized at 20 kHz and
measured by slope (10–50 % of fEPSP rising phase). The expressed BlgC in DG granule cells
was photoactivated using a blue LED light (1.5 mW, 5 min) under an objective lens (4X,
NA0.1). Tetanus was induced with a bipolar tetanic stimulation (100 Hz, 0.15 msec pulses
delivered in 4 trains of 0.5 sec duration, 20 sec apart) for the perforant path stimulation. In the
case of photoactivation of BlgC with tetanic stimulation experiments, tetanic stimulation was
induced after the first minute of light exposure during photoactivation. In the time course
experiments, field responses were plotted by normalizing to the baseline fEPSP slope (average
of the 10-minute period prior to tetanic stimulation).
3.10 Statistical analysis All statistical analysis was performed using KyPlot. Data are presented as mean ± SEM. Steel-
Dunnett non parametric test was used for all structural plasticity comparisons; n is defined as
number of dendritic spines (per # of neurons is indicated in figure legends). ANOVA was used
108
for all mouse behavior comparisons, as indicated in figure legends. Differences were
considered statistically significant at p > 0.05.
3.11 Mouse behavior Mice were surgically equipped with either a bilateral fiber optic LED system (n=14) or, as a
control, a “dummy” optic fiber, which produces no light but exhibits the same physical
properties (n=12) and housed in a Home Cage Monitoring System with automated video
tracking and habituated to wearing a detachable battery (necessary to power the LED). The
battery (or dummy battery) was attached and turned on to produce cGMP during each test
(light-activation details available in each figure). All behavior tests were performed as
previously described [227, 228]. The control and cGMP mouse cohorts were counterbalanced
such that each test was performed on equal number of control and cGMP mice simultaneously
(in the case of multiple chambers). In single chamber tests, a control mouse test was followed
by cGMP mouse test until all mice were evaluated. After the behavior testing was complete,
all mice were sacrificed and checked for the expression of BlgC, which resulted in the
exclusion of 1 BlgC-cGMP and 2 dummy mice from the results pool.
3.11.1 General Health and Neurological Screen (GHNS)
The general health and neurological screen was conducted as previously described [227]. Body
weight and rectal temperature were measured. Neuromuscular strength was assessed using the
grip strength and wire hang tests. A grip strength meter (O’Hara & Co., Tokyo, Japan) was
used to assess forelimb grip strength. Mice were lifted and held by their tail so that their
forepaws could grasp a wire grid. The mice were then gently pulled backward by the tail until
they released the grid. The peak force applied by the forelimbs of the mouse was recorded in
Newtons (N). Each mouse was tested three times, and the largest value was used for statistical
109
analysis. In the wire hang test, the mouse was placed on a wire mesh that was then inverted,
and the latency to fall from the wire was recorded with a 60 s cut-off time. The hot plate test
was performed to evaluate sensitivity to painful stimuli. Mice were placed on a 55.0 ± 0.3 ° C
hot plate (Columbus Instruments, Columbus, OH, USA), and latency to the first paw response
was recorded with a 15 s cut-off time. The paw response was defined as either a foot shake or
a paw lick. Motor coordination and balance were tested with a rotarod test. The rotarod test,
(UGO Basile Accelerating Rotarod, Varese, Italy), was performed by placing mice on
accelerating rotating drums (3 cm diameter) and measuring the time each animal was able to
maintain its balance on the rod. The speed of the rotarod accelerated from 4 to 40 rpm over a
5-min period. All mice were fitted with a battery for 5 minutes immediately prior to the
screening with the light turned on to generate cGMP in BlgC mice.
3.11.2 Light/Dark test (LD)
The light/dark transition test apparatus consisted of a cage (21 × 42 × 25 cm) divided into two
sections of equal size by a partition with a door (O’Hara & Co., Tokyo, Japan). One chamber
was brightly illuminated (390 lux), whereas the other chamber was dark (2 lux). Mice were
placed into the dark chamber and allowed to move freely between the chambers with the door
open for 10 min. The total number of transitions between chambers, time spent in each chamber
(s), latency to first enter the light chamber (s), and distance traveled in each chamber (cm) were
recorded automatically by ImageLD software. All mice were fitted with a battery for 5 minutes
immediately prior to the test with the light turned on to generate cGMP in BlgC mice.
3.11.3 Elevated Plus maze (EP)
The elevated plus maze consisted of two open arms (25 × 5 cm, with 3-mm-high ledges) and
two closed arms (25 × 5 cm, with 15-cm-high transparent walls) of the same size (O’Hara &
110
Co., Tokyo, Japan). The arms and central square were made of white plastic plates and were
elevated to a height of 55 cm above the floor. The arms of the same type were arranged at
opposite sides to each other. The center of the maze was illuminated at 100 lux. Each mouse
was placed in the central square of the maze (5 × 5 cm), facing one of the closed arms, and was
recorded for 10 min. The distance traveled (cm), number of total entries into arms, percentage
of entries into open arms, and percentage of time spent in open arms were calculated
automatically using ImageEP software. All mice were fitted with a battery during the test with
the light turned on for the first 5 minutes to generate cGMP in BlgC mice.
3.11.4 Open Field test (OF)
The open field test apparatus was a transparent square cage (42 × 42 × 30 cm; Accuscan
Instruments, Columbus, OH, USA). The center of the floor was illuminated at 100 lux. Each
mouse was placed in the open field apparatus and recorded for 120 min. Total distance traveled
(cm), and time spent in the center area (20 × 20 cm), were measured. All mice were fitted with
a battery for 10 minutes immediately prior to the test and light turned on to generate cGMP in
BlgC mice.
3.11.5 Crawley’s Social Interaction test (CSI)
The CSI test apparatus had three chambers (20 × 40 × 22 cm) separated by two transparent
partitions each with an opening (5 × 3 cm), and a lid with an infrared CCD camera. A male
mouse (8–12 weeks old C57BL/6J, termed Stranger 1) that had no prior contact with the subject
mice was enclosed in a cylinder cage (9 cm ϕ, set in the left chamber) that allowed nose
contacts. The subject mouse was released in the middle chamber and allowed to explore for
10 min, while the time spent in each chamber and within 5 cm from each cage was measured
automatically using ImageCSI software. Subsequently, another unfamiliar mouse (Stranger 2)
111
was placed in another cylinder cage (in the right chamber) and monitored likewise for another
10 min. All mice were fitted with a battery for 5 minutes immediately prior to the test with the
light turned on to generate cGMP in BlgC mice.
3.11.6 Social Interaction test (SI)
Two mice from the cohort (paired as BlgC and control, whenever possible), were placed into
a box together (40 × 40 × 30 cm; O’Hara & Co., Tokyo, Japan) and were allowed to explore
freely for 10 min. Mouse behavior was analyzed automatically using ImageSI software. The
total duration of contacts (s), number of contacts, total duration of active contacts (s), mean
duration per contact, and total distance traveled (cm) were measured. The active contact was
defined as follows: images were captured at three frames per second, and distance traveled
between two successive frames was calculated for each mouse. If the two mice contacted each
other and the distance traveled by either mouse was 5 cm and more, the behavior was
considered an “active contact.” All mice were fitted with a battery during the test with the light
turned on for the first 5 minutes to generate cGMP in BlgC mice.
3.11.7 Object Localization test (OLT)
During familiarization phase, the mice were placed in a solid plastic chamber (30 × 30 × 30
cm) containing two identical objects, equidistant from the center and walls, placed in opposite
corners. After 5 minutes, mice were placed back into their home cage and then returned to the
same chamber where one of the objects was replaced with a novel object (in the same location).
After 5 minutes of this test phase, the mice were placed back into their home cage and then
returned to the same chamber where one of the objects was replaced with a new novel object
placed in a novel location (5 minutes). Objects of four types (a black sphere, a black cone, a
112
column with vertical stripes and a cube with black dots) were used and data for the time around
each object (sec) and number of entries (based on the selected region of interest) was collected.
Data acquisition and analysis were performed automatically using ImageOF. All mice were
fitted with a battery for 5 minutes during familiarization phase with the light turned on to
generate cGMP in BlgC mice.
3.11.8 Barnes Maze (BM)
The Barnes circular maze task was conducted on a white circular surface (1.0 m in diameter,
with 12 holes equally spaced around the perimeter; O’hara & Co., Tokyo, Japan). The circular
open field was elevated 75 cm from the floor. A black Plexiglas escape box (17 × 13 × 7 cm),
which had paper cage bedding on its bottom, was located under one of the holes. The hole
above the escape box represented the target, analogous to the hidden platform in the Morris
water maze task. The location of the target was consistent for a given mouse but randomized
across mice. The maze was rotated daily, with the spatial location of the target unchanged with
respect to the distal visual room cues, to prevent a bias based on olfactory or proximal cues
within the maze. Mice were trained for 15 trials, with one to four training trials per day. The
number of errors, latency to reach the target (s), and distance traveled to reach the target (cm)
were automatically calculated by ImageBM software.
24 hours after the last training, a probe trial (PT1) was conducted without the escape box for
3 min, to confirm that this spatial task was acquired based on navigation by distal environment
cues. The time spent around each hole, number of errors, moving time (s), distance traveled
(cm), and moving speed (cm/s) were recorded using ImageBM software.
1 month after the PT1, another probe trial (PT2) was performed to evaluate memory retention.
After that, mice were subjected to a reversal training session. In reversal training, the location
113
of the target hole was changed to the opposite side of the maze, and the mice were trained for
9 trials. 24 hours after the last reversal training, a probe trial (PT3) was performed. Data
acquisition and analysis were performed automatically using ImageBM. All mice were fitted
with a battery for 5 minutes immediately prior to each training session (original training and
reversal training) with the light was turned on to generate cGMP in BlgC mice.
3.11.9 T-Maze (Forced Alternation) (TMFA)
The forced alternation task was conducted using an automatic T-maze (O'Hara & Co., Tokyo,
Japan) constructed of white plastics runways with walls 25-cm high. The maze was partitioned
off into 6 areas by sliding doors that can be opened downward. The stem of T was composed
of area S2 (13 × 24 cm) and the arms of T were composed of area A1 and A2 (11.5 × 20.5 cm).
Area P1 and P2 were the connecting passage way from the arm (area A1 or A2) to the start
compartment (area S1). The end of each arm was equipped with a pellet dispenser that could
provide food reward. The pellet sensors were able to record automatically pellet intake by the
mice. One week before the pre-training, mice were deprived of food until their body weight
was reduced to 80–85% of the initial level. Mice were kept on a maintenance diet throughout
the course of all the T-maze experiments. Before the first trial, mice were subjected to three
10-min adaptation sessions, during which they were allowed to freely explore the T-maze with
all doors open and both arms baited with food. One day after the adaptation session, mice were
subjected to a forced alternation protocol for 5 sessions (one session consisting of 10 trials per
day; cutoff time, 50 min). On the first (sample) trial of each pair, the mouse was forced to
choose one of the arms of the T (area A1 or A2), and received the reward at the end of the arm.
Choosing the incorrect arm resulted in no reward and confinement to the arm for 10 sec. After
the mouse consumed the pellet or the mouse stayed more than 10 sec without consuming the
114
pellet, door that separated the arm (area A1 or A2) and connecting passage way (area P1 or P2)
would be opened and the mouse could return to the starting compartment (area S1), via
connecting passage way, by itself. The mouse was then given a 3 sec delay and a free choice
between both T arms and rewarded for choosing the other arm that was not chosen on the first
trial of the pair. The location of the sample arm (left or right) was varied pseudo-randomly
across trials using Gellermann schedule so that mice received equal numbers of left and right
presentations. On the 16–21st day, delay (10, 30 or 60 sec) was applied after the sample trial.
Data acquisition, control of sliding doors, and data analysis were performed by Image TM
software. All mice were fitted with a battery for 5 minutes immediately prior to each training
session (consisting of 10 L/R trials) with the light was turned on to generate cGMP in BlgC
mice.
115
References 1. Gray, E.G., Electron microscopy of synaptic contacts on dendrite spines of the
cerebral cortex. Nature, 1959. 183(4675): p. 1592-3.
2. Bliss, T.V. and T. Lomo, Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J Physiol, 1973. 232(2): p. 331-56.
3. Mulcahy, B., et al., A Pipeline for Volume Electron Microscopy of the Caenorhabditis elegans Nervous System. Front Neural Circuits, 2018. 12: p. 94.
4. White, J.G., et al., The structure of the nervous system of the nematode Caenorhabditis elegans. Philos Trans R Soc Lond B Biol Sci, 1986. 314(1165): p. 1-340.
5. Leiss, F., et al., Characterization of dendritic spines in the Drosophila central nervous system. Dev Neurobiol, 2009. 69(4): p. 221-34.
6. Coss, R.G., J.G. Brandon, and A. Globus, Changes in morphology of dendritic spines on honeybee calycal interneurons associated with cumulative nursing and foraging experiences. Brain Res, 1980. 192(1): p. 49-59.
7. Burgess, J.W. and R.G. Coss, Crowded jewel fish show changes in dendritic spine density and spine morphology. Neurosci Lett, 1980. 17(3): p. 277-81.
8. Gilbert, M.T. and K. Soderstrom, Novel song-stimulated dendritic spine formation and Arc/Arg3.1 expression in zebra finch auditory telencephalon are disrupted by cannabinoid agonism. Brain Res, 2013. 1541: p. 9-21.
9. Mancuso, J.J., et al., Methods of dendritic spine detection: from Golgi to high-resolution optical imaging. Neuroscience, 2013. 251: p. 129-40.
10. Yuste, R., The discovery of dendritic spines by Cajal. Front Neuroanat, 2015. 9: p. 18.
11. Zito, K. and V.N. Murthy, Dendritic spines. Curr Biol, 2002. 12(1): p. R5.
12. Ziv, N.E. and S.J. Smith, Evidence for a role of dendritic filopodia in synaptogenesis and spine formation. Neuron, 1996. 17(1): p. 91-102.
13. Harris, K.M., F.E. Jensen, and B. Tsao, Three-dimensional structure of dendritic spines and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: implications for the maturation of synaptic physiology and long-term potentiation. J Neurosci, 1992. 12(7): p. 2685-705.
14. Mahmmoud, R.R., et al., Spatial and Working Memory Is Linked to Spine Density and Mushroom Spines. PLoS One, 2015. 10(10): p. e0139739.
116
15. Gallinaro, J.V. and S. Rotter, Associative properties of structural plasticity based on
firing rate homeostasis in recurrent neuronal networks. Sci Rep, 2018. 8(1): p. 3754.
16. Caroni, P., F. Donato, and D. Muller, Structural plasticity upon learning: regulation and functions. Nat Rev Neurosci, 2012. 13(7): p. 478-90.
17. Bailey, C.H. and M. Chen, Structural plasticity at identified synapses during long-term memory in Aplysia. J Neurobiol, 1989. 20(5): p. 356-72.
18. Van Harreveld, A. and E. Fifkova, Swelling of dendritic spines in the fascia dentata after stimulation of the perforant fibers as a mechanism of post-tetanic potentiation. Exp Neurol, 1975. 49(3): p. 736-49.
19. Fischer, M., et al., Rapid actin-based plasticity in dendritic spines. Neuron, 1998. 20(5): p. 847-54.
20. Svoboda, K. and R. Yasuda, Principles of two-photon excitation microscopy and its applications to neuroscience. Neuron, 2006. 50(6): p. 823-39.
21. Zhou, Q., K.J. Homma, and M.M. Poo, Shrinkage of dendritic spines associated with long-term depression of hippocampal synapses. Neuron, 2004. 44(5): p. 749-57.
22. Okamoto, K., et al., Rapid and persistent modulation of actin dynamics regulates postsynaptic reorganization underlying bidirectional plasticity. Nat Neurosci, 2004. 7(10): p. 1104-12.
23. Dominguez, R. and K.C. Holmes, Actin structure and function. Annu Rev Biophys, 2011. 40: p. 169-86.
24. Hotulainen, P. and C.C. Hoogenraad, Actin in dendritic spines: connecting dynamics to function. J Cell Biol, 2010. 189(4): p. 619-29.
25. Pontrello, C.G. and I.M. Ethell, Accelerators, Brakes, and Gears of Actin Dynamics in Dendritic Spines. Open Neurosci J, 2009. 3: p. 67-86.
26. Tada, T. and M. Sheng, Molecular mechanisms of dendritic spine morphogenesis. Curr Opin Neurobiol, 2006. 16(1): p. 95-101.
27. Cingolani, L.A. and Y. Goda, Actin in action: the interplay between the actin cytoskeleton and synaptic efficacy. Nat Rev Neurosci, 2008. 9(5): p. 344-56.
28. Dillon, C. and Y. Goda, The actin cytoskeleton: integrating form and function at the synapse. Annu Rev Neurosci, 2005. 28: p. 25-55.
29. Okamoto, K. and Y. Hayashi, Visualization of F-actin and G-actin equilibrium using fluorescence resonance energy transfer (FRET) in cultured cells and neurons in slices. Nat Protoc, 2006. 1(2): p. 911-9.
117
30. Honkura, N., et al., The subspine organization of actin fibers regulates the structure
and plasticity of dendritic spines. Neuron, 2008. 57(5): p. 719-29.
31. Bosch, M., et al., Structural and molecular remodeling of dendritic spine substructures during long-term potentiation. Neuron, 2014. 82(2): p. 444-59.
32. Ichetovkin, I., W. Grant, and J. Condeelis, Cofilin produces newly polymerized actin filaments that are preferred for dendritic nucleation by the Arp2/3 complex. Curr Biol, 2002. 12(1): p. 79-84.
33. Korobova, F. and T. Svitkina, Molecular architecture of synaptic actin cytoskeleton in hippocampal neurons reveals a mechanism of dendritic spine morphogenesis. Mol Biol Cell, 2010. 21(1): p. 165-76.
34. Djinovic-Carugo, K., et al., Structure of the alpha-actinin rod: molecular basis for cross-linking of actin filaments. Cell, 1999. 98(4): p. 537-46.
35. Koganezawa, N., et al., The role of drebrin in dendritic spines. Mol Cell Neurosci, 2017. 84: p. 85-92.
36. Okamoto, K., et al., The role of CaMKII as an F-actin-bundling protein crucial for maintenance of dendritic spine structure. Proc Natl Acad Sci U S A, 2007. 104(15): p. 6418-23.
37. Sanabria, H., et al., {beta}CaMKII regulates actin assembly and structure. J Biol Chem, 2009. 284(15): p. 9770-80.
38. Sjoblom, B., A. Salmazo, and K. Djinovic-Carugo, Alpha-actinin structure and regulation. Cell Mol Life Sci, 2008. 65(17): p. 2688-701.
39. Hayashi, K., et al., Modulatory role of drebrin on the cytoskeleton within dendritic spines in the rat cerebral cortex. J Neurosci, 1996. 16(22): p. 7161-70.
40. Hodges, J.L., et al., alpha-Actinin-2 mediates spine morphology and assembly of the post-synaptic density in hippocampal neurons. PLoS One, 2014. 9(7): p. e101770.
41. Yoshimura, Y., et al., Molecular constituents of the postsynaptic density fraction revealed by proteomic analysis using multidimensional liquid chromatography-tandem mass spectrometry. J Neurochem, 2004. 88(3): p. 759-68.
42. Van Troys, M., et al., Ins and outs of ADF/cofilin activity and regulation. Eur J Cell Biol, 2008. 87(8-9): p. 649-67.
43. Bamburg, J.R. and B.W. Bernstein, Roles of ADF/cofilin in actin polymerization and beyond. F1000 Biol Rep, 2010. 2: p. 62.
44. Erondu, N.E. and M.B. Kennedy, Regional distribution of type II Ca2+/calmodulin-dependent protein kinase in rat brain. J Neurosci, 1985. 5(12): p. 3270-7.
118
45. Kim, K., et al., A Temporary Gating of Actin Remodeling during Synaptic Plasticity
Consists of the Interplay between the Kinase and Structural Functions of CaMKII. Neuron, 2015. 87(4): p. 813-26.
46. Lee, S.J., et al., Activation of CaMKII in single dendritic spines during long-term potentiation. Nature, 2009. 458(7236): p. 299-304.
47. Lisman, J., H. Schulman, and H. Cline, The molecular basis of CaMKII function in synaptic and behavioural memory. Nat Rev Neurosci, 2002. 3(3): p. 175-90.
48. Shen, K., et al., CaMKIIbeta functions as an F-actin targeting module that localizes CaMKIIalpha/beta heterooligomers to dendritic spines. Neuron, 1998. 21(3): p. 593-606.
49. Hanson, P.I. and H. Schulman, Inhibitory autophosphorylation of multifunctional Ca2+/calmodulin-dependent protein kinase analyzed by site-directed mutagenesis. J Biol Chem, 1992. 267(24): p. 17216-24.
50. Shields, S.M., T.S. Ingebritsen, and P.T. Kelly, Identification of protein phosphatase 1 in synaptic junctions: dephosphorylation of endogenous calmodulin-dependent kinase II and synapse-enriched phosphoproteins. J Neurosci, 1985. 5(12): p. 3414-22.
51. Strack, S., et al., Differential inactivation of postsynaptic density-associated and soluble Ca2+/calmodulin-dependent protein kinase II by protein phosphatases 1 and 2A. J Neurochem, 1997. 68(5): p. 2119-28.
52. Beavo, J.A. and L.L. Brunton, Cyclic nucleotide research -- still expanding after half a century. Nat Rev Mol Cell Biol, 2002. 3(9): p. 710-8.
53. Chetkovich, D.M., et al., N-methyl-D-aspartate receptor activation increases cAMP levels and voltage-gated Ca2+ channel activity in area CA1 of hippocampus. Proc Natl Acad Sci U S A, 1991. 88(15): p. 6467-71.
54. Chetkovich, D.M. and J.D. Sweatt, nMDA receptor activation increases cyclic AMP in area CA1 of the hippocampus via calcium/calmodulin stimulation of adenylyl cyclase. J Neurochem, 1993. 61(5): p. 1933-42.
55. Conn, P.J. and J.P. Pin, Pharmacology and functions of metabotropic glutamate receptors. Annual Review of Pharmacology and Toxicology, 1997. 37(Journal Article): p. 205-237.
56. Gilman, A.G., G-Proteins - Transducers of Receptor-Generated Signals. Annual Review of Biochemistry, 1987. 56(Journal Article): p. 615-649.
57. Seeman, P., Brain Dopamine-Receptors. Pharmacological reviews, 1980. 32(3): p. 229-313.
119
58. Huang, Y.Y., X.C. Li, and E.R. Kandel, cAMP contributes to mossy fiber LTP by
initiating both a covalently mediated early phase and macromolecular synthesis-dependent late phase. Cell, 1994. 79(1): p. 69-79.
59. Frey, U., Y.Y. Huang, and E.R. Kandel, Effects of cAMP simulate a late stage of LTP in hippocampal CA1 neurons. Science, 1993. 260(5114): p. 1661-4.
60. Bolshakov, V.Y., et al., Recruitment of new sites of synaptic transmission during the cAMP-dependent late phase of LTP at CA3-CA1 synapses in the hippocampus. Neuron, 1997. 19(3): p. 635-51.
61. Rehmann, H., A. Wittinghofer, and J.L. Bos, Capturing cyclic nucleotides in action: snapshots from crystallographic studies. Nat Rev Mol Cell Biol, 2007. 8(1): p. 63-73.
62. Scott Bitner, R., Cyclic AMP response element-binding protein (CREB) phosphorylation: a mechanistic marker in the development of memory enhancing Alzheimer's disease therapeutics. Biochem Pharmacol, 2012. 83(6): p. 705-14.
63. Govindarajan, A., et al., The Dendritic Branch Is the Preferred Integrative Unit for Protein Synthesis-Dependent LTP. Neuron, 2011. 69(1): p. 132-146.
64. Otmakhov, N., et al., Persistent accumulation of calcium/calmodulin-dependent protein kinase II in dendritic spines after induction of NMDA receptor-dependent chemical long-term potentiation. J Neurosci, 2004. 24(42): p. 9324-31.
65. Luyben, T., et al. Rapid postsynaptic cAMP signaling regulates structural and functional potentiation underlying learning and memory. in Neuroscience meeting. 2016. San Diego, CA: Society for Neuroscience.
66. Borovac, J., M. Bosch, and K. Okamoto, Regulation of actin dynamics during structural plasticity of dendritic spines: Signaling messengers and actin-binding proteins. Mol Cell Neurosci, 2018.
67. Blitzer, R.D., et al., Gating of CaMKII by cAMP-regulated protein phosphatase activity during LTP. Science, 1998. 280(5371): p. 1940-2.
68. de Vente, J., et al., Nitric-oxide-induced cGMP synthesis in cholinergic neurons in the rat brain. Exp Brain Res, 2001. 136(4): p. 480-91.
69. Chalimoniuk, M., B. Glod, and J. Strosznajder, NMDA receptor mediated nitric oxide dependent cGMP synthesis in brain cortex and hippocampus. Effect of ischemia on NO related biochemical processes during reperfusion. Neurol Neurochir Pol, 1996. 30 Suppl 2: p. 65-84.
70. Altar, C.A., W.C. Boyar, and H.S. Kim, Discriminatory roles for D1 and D2 dopamine receptor subtypes in the in vivo control of neostriatal cyclic GMP. Eur J Pharmacol, 1990. 181(1-2): p. 17-21.
120
71. Mayer, B., M. John, and E. Bohme, Purification of a Ca2+/calmodulin-dependent
nitric oxide synthase from porcine cerebellum. Cofactor-role of tetrahydrobiopterin. FEBS Lett, 1990. 277(1-2): p. 215-9.
72. Mayer, B., et al., Regulation of neuronal nitric oxide and cyclic GMP formation by Ca2+. J Neurochem, 1992. 59(6): p. 2024-9.
73. Murad, F., et al., Guanylate cyclase: activation by azide, nitro compounds, nitric oxide, and hydroxyl radical and inhibition by hemoglobin and myoglobin. Adv Cyclic Nucleotide Res, 1978. 9: p. 145-58.
74. Francis, S.H., et al., cGMP-dependent protein kinases and cGMP phosphodiesterases in nitric oxide and cGMP action. Pharmacol Rev, 2010. 62(3): p. 525-63.
75. Martins, T.J., M.C. Mumby, and J.A. Beavo, Purification and characterization of a cyclic GMP-stimulated cyclic nucleotide phosphodiesterase from bovine tissues. J Biol Chem, 1982. 257(4): p. 1973-9.
76. Kuzmiski, J.B. and B.A. MacVicar, Cyclic nucleotide-gated channels contribute to the cholinergic plateau potential in hippocampal CA1 pyramidal neurons. J Neurosci, 2001. 21(22): p. 8707-14.
77. Van Staveren, W.C., et al., mRNA expression patterns of the cGMP-hydrolyzing phosphodiesterases types 2, 5, and 9 during development of the rat brain. J Comp Neurol, 2003. 467(4): p. 566-80.
78. Bender, A.T. and J.A. Beavo, Cyclic nucleotide phosphodiesterases: molecular regulation to clinical use. Pharmacol Rev, 2006. 58(3): p. 488-520.
79. Nikonenko, I., et al., Nitric oxide mediates local activity-dependent excitatory synapse development. Proc Natl Acad Sci U S A, 2013. 110(44): p. E4142-51.
80. Wang, H.G., et al., Presynaptic and postsynaptic roles of NO, cGK, and RhoA in long-lasting potentiation and aggregation of synaptic proteins. Neuron, 2005. 45(3): p. 389-403.
81. Applewhite, D.A., et al., Ena/VASP proteins have an anti-capping independent function in filopodia formation. Mol Biol Cell, 2007. 18(7): p. 2579-91.
82. Zhuang, S., et al., Vasodilator-stimulated phosphoprotein activation of serum-response element-dependent transcription occurs downstream of RhoA and is inhibited by cGMP-dependent protein kinase phosphorylation. J Biol Chem, 2004. 279(11): p. 10397-407.
83. Gudi, T., et al., cGMP-dependent protein kinase inhibits serum-response element-dependent transcription by inhibiting rho activation and functions. J Biol Chem, 2002. 277(40): p. 37382-93.
121
84. Lorente, d.N., Studies on the structure of the cerebral cortex. II: Continuation of the
study of the ammonic system. J Psychol Neurol, 1934. 46: p. 113–177.
85. Petrantonakis, P.C. and P. Poirazi, A compressed sensing perspective of hippocampal function. Front Syst Neurosci, 2014. 8: p. 141.
86. Basu, J. and S.A. Siegelbaum, The Corticohippocampal Circuit, Synaptic Plasticity, and Memory. Cold Spring Harb Perspect Biol, 2015. 7(11).
87. Bliss, T.V. and A.R. Gardner-Medwin, Long-lasting potentiation of synaptic transmission in the dentate area of the unanaestetized rabbit following stimulation of the perforant path. J Physiol, 1973. 232(2): p. 357-74.
88. Bliss, T.V. and G.L. Collingridge, A synaptic model of memory: long-term potentiation in the hippocampus. Nature, 1993. 361(6407): p. 31-9.
89. Bukalo, O. and A. Dityatev, Analysis of neural cell functions in gene knockout mice: electrophysiological investigation of synaptic plasticity in acute hippocampal slices. Methods Enzymol, 2006. 417: p. 52-66.
90. Kochlamazashvili G., S.O., Dityatev A., Extracellular Recordings of Synaptic Plasticity and Network Oscillations in Hippocampal Slices. Neuromethods, ed. H.M. Fellin T. Vol. 67. 2011.
91. Nicoll, R.A. and K.W. Roche, Long-term potentiation: peeling the onion. Neuropharmacology, 2013. 74: p. 18-22.
92. Huang, E.P., Synaptic plasticity: going through phases with LTP. Curr Biol, 1998. 8(10): p. R350-2.
93. Smolen, P., D.A. Baxter, and J.H. Byrne, A model of the roles of essential kinases in the induction and expression of late long-term potentiation. Biophys J, 2006. 90(8): p. 2760-75.
94. Chao, L.H., et al., Intersubunit capture of regulatory segments is a component of cooperative CaMKII activation. Nat Struct Mol Biol, 2010. 17(3): p. 264-72.
95. Mullasseril, P., et al., A structural mechanism for maintaining the 'on-state' of the CaMKII memory switch in the post-synaptic density. J Neurochem, 2007. 103(1): p. 357-64.
96. Lisman, J.E., A mechanism for memory storage insensitive to molecular turnover: a bistable autophosphorylating kinase. Proc Natl Acad Sci U S A, 1985. 82(9): p. 3055-7.
97. Hanse, E. and B. Gustafsson, Postsynaptic, but not presynaptic, activity controls the early time course of long-term potentiation in the dentate gyrus. J Neurosci, 1992. 12(8): p. 3226-40.
122
98. Maren, S., et al., Postsynaptic factors in the expression of long-term potentiation
(LTP): increased glutamate receptor binding following LTP induction in vivo. Proc Natl Acad Sci U S A, 1993. 90(20): p. 9654-8.
99. Nguyen, P.V. and E.R. Kandel, A macromolecular synthesis-dependent late phase of long-term potentiation requiring cAMP in the medial perforant pathway of rat hippocampal slices. J Neurosci, 1996. 16(10): p. 3189-98.
100. Wu, J., M.J. Rowan, and R. Anwyl, Long-term potentiation is mediated by multiple kinase cascades involving CaMKII or either PKA or p42/44 MAPK in the adult rat dentate gyrus in vitro. J Neurophysiol, 2006. 95(6): p. 3519-27.
101. Steward, O., F. Huang, and J.F. Guzowski, A form of perforant path LTP can occur without ERK1/2 phosphorylation or immediate early gene induction. Learn Mem, 2007. 14(6): p. 433-45.
102. Schulz, S., et al., Direct evidence for biphasic cAMP responsive element-binding protein phosphorylation during long-term potentiation in the rat dentate gyrus in vivo. J Neurosci, 1999. 19(13): p. 5683-92.
103. Liu, X., et al., Identification and Manipulation of Memory Engram Cells. Cold Spring Harb Symp Quant Biol, 2014. 79: p. 59-65.
104. Wang, W., et al., A Primary Cortical Input to Hippocampus Expresses a Pathway-Specific and Endocannabinoid-Dependent Form of Long-Term Potentiation. eNeuro, 2016. 3(4).
105. Henderson, J.T., et al., The receptor tyrosine kinase EphB2 regulates NMDA-dependent synaptic function. Neuron, 2001. 32(6): p. 1041-56.
106. Saab, B.J., et al., NCS-1 in the dentate gyrus promotes exploration, synaptic plasticity, and rapid acquisition of spatial memory. Neuron, 2009. 63(5): p. 643-56.
107. Wilson, D.I., et al., Lateral entorhinal cortex is necessary for associative but not nonassociative recognition memory. Hippocampus, 2013. 23(12): p. 1280-90.
108. Deisseroth, K., H. Bito, and R.W. Tsien, Signaling from synapse to nucleus: Postsynaptic CREB phosphorylation during multiple forms of hippocampal synaptic plasticity. Neuron, 1996. 16(1): p. 89-101.
109. Berkowitz, L.A., K.T. Riabowol, and M.Z. Gilman, Multiple Sequence Elements of a Single Functional Class are Required for Cyclic-Amp Responsiveness of the Mouse C-Fos Promoter. Molecular and cellular biology, 1989. 9(10): p. 4272-4281.
110. Walton, M., et al., Immediate early gene transcription and synaptic modulation. Journal of neuroscience research, 1999. 58(1): p. 96-106.
111. Kleppisch, T., et al., Hippocampal cGMP-dependent protein kinase I supports an age- and protein synthesis-dependent component of long-term potentiation but is not
123
essential for spatial reference and contextual memory. J Neurosci, 2003. 23(14): p. 6005-12.
112. Boulton, C.L., E. Southam, and J. Garthwaite, Nitric oxide-dependent long-term potentiation is blocked by a specific inhibitor of soluble guanylyl cyclase. Neuroscience, 1995. 69(3): p. 699-703.
113. Zhuo, M., et al., Role of guanylyl cyclase and cGMP-dependent protein kinase in long-term potentiation. Nature, 1994. 368(6472): p. 635-9.
114. Haley, J.E., G.L. Wilcox, and P.F. Chapman, The role of nitric oxide in hippocampal long-term potentiation. Neuron, 1992. 8(2): p. 211-6.
115. Son, H., et al., The specific role of cGMP in hippocampal LTP. Learn Mem, 1998. 5(3): p. 231-45.
116. Feil, R., et al., Impairment of LTD and cerebellar learning by Purkinje cell-specific ablation of cGMP-dependent protein kinase I. J Cell Biol, 2003. 163(2): p. 295-302.
117. Hartell, N.A., cGMP acts within cerebellar Purkinje cells to produce long term depression via mechanisms involving PKC and PKG. Neuroreport, 1994. 5(7): p. 833-6.
118. Hartell, N.A., Inhibition of cGMP breakdown promotes the induction of cerebellar long-term depression. J Neurosci, 1996. 16(9): p. 2881-90.
119. Blackwell, K.T. and J. Jedrzejewska-Szmek, Molecular mechanisms underlying neuronal synaptic plasticity: systems biology meets computational neuroscience in the wilds of synaptic plasticity. Wiley Interdiscip Rev Syst Biol Med, 2013. 5(6): p. 717-31.
120. Micheva, K.D., et al., Retrograde regulation of synaptic vesicle endocytosis and recycling. Nat Neurosci, 2003. 6(9): p. 925-32.
121. Kleppisch, T. and R. Feil, cGMP signalling in the mammalian brain: role in synaptic plasticity and behaviour. Handb Exp Pharmacol, 2009(191): p. 549-79.
122. Stricker, S. and D. Manahan-Vaughan, Regulation of long-term depression by increases in [guanosine 3',5'-cyclic monophosphate] in the hippocampal CA1 region of freely behaving rats. Neuroscience, 2009. 158(1): p. 159-66.
123. Serulle, Y., et al., A GluR1-cGKII interaction regulates AMPA receptor trafficking. Neuron, 2007. 56(4): p. 670-88.
124. Abel, T. and K.M. Lattal, Molecular mechanisms of memory acquisition, consolidation and retrieval. Curr Opin Neurobiol, 2001. 11(2): p. 180-7.
125. Squire, L.R. and P. Alvarez, Retrograde amnesia and memory consolidation: a neurobiological perspective. Curr Opin Neurobiol, 1995. 5(2): p. 169-77.
124
126. Squire, L.R., et al., Memory consolidation. Cold Spring Harb Perspect Biol, 2015.
7(8): p. a021766.
127. Schacter, D.L. and C.S. Dodson, Misattribution, false recognition and the sins of memory. Philos Trans R Soc Lond B Biol Sci, 2001. 356(1413): p. 1385-93.
128. Nader, K. and E.O. Einarsson, Memory reconsolidation: an update. Ann N Y Acad Sci, 2010. 1191: p. 27-41.
129. Greenberg, D.L. and M. Verfaellie, Interdependence of episodic and semantic memory: evidence from neuropsychology. J Int Neuropsychol Soc, 2010. 16(5): p. 748-53.
130. Tulving, E., Episodic memory: from mind to brain. Annu Rev Psychol, 2002. 53: p. 1-25.
131. Pause, B.M., et al., Perspectives on episodic-like and episodic memory. Front Behav Neurosci, 2013. 7: p. 33.
132. Baddeley, A., Working memory. Science, 1992. 255(5044): p. 556-9.
133. Paul, C.M., G. Magda, and S. Abel, Spatial memory: Theoretical basis and comparative review on experimental methods in rodents. Behav Brain Res, 2009. 203(2): p. 151-64.
134. Olton, D.S., Mazes, maps, and memory. Am Psychol, 1979. 34(7): p. 583-96.
135. Kohler, S., M. Moscovitch, and B. Melo, Episodic memory for object location versus episodic memory for object identity: do they rely on distinct encoding processes? Mem Cognit, 2001. 29(7): p. 948-59.
136. Lee, S.A., et al., Working memory and reference memory tests of spatial navigation in mice (Mus musculus). J Comp Psychol, 2015. 129(2): p. 189-97.
137. Rosenfeld, C.S. and S.A. Ferguson, Barnes maze testing strategies with small and large rodent models. J Vis Exp, 2014(84): p. e51194.
138. Shoji, H., et al., T-maze forced alternation and left-right discrimination tasks for assessing working and reference memory in mice. J Vis Exp, 2012(60).
139. Baumann, O., A.J. Skilleter, and J.B. Mattingley, Short-term memory maintenance of object locations during active navigation: which working memory subsystem is essential? PLoS One, 2011. 6(5): p. e19707.
140. Sutherland, R.J., I.Q. Whishaw, and B. Kolb, A behavioural analysis of spatial localization following electrolytic, kainate- or colchicine-induced damage to the hippocampal formation in the rat. Behav Brain Res, 1983. 7(2): p. 133-53.
125
141. McNaughton, B.L., et al., Hippocampal granule cells are necessary for normal
spatial learning but not for spatially-selective pyramidal cell discharge. Exp Brain Res, 1989. 76(3): p. 485-96.
142. Xavier, G.F., F.J. Oliveira-Filho, and A.M. Santos, Dentate gyrus-selective colchicine lesion and disruption of performance in spatial tasks: difficulties in "place strategy" because of a lack of flexibility in the use of environmental cues? Hippocampus, 1999. 9(6): p. 668-81.
143. Ramirez, S., et al., Creating a false memory in the hippocampus. Science, 2013. 341(6144): p. 387-91.
144. Redondo, R.L., et al., Bidirectional switch of the valence associated with a hippocampal contextual memory engram. Nature, 2014. 513(7518): p. 426-30.
145. Ramirez, S., et al., Activating positive memory engrams suppresses depression-like behaviour. Nature, 2015. 522(7556): p. 335-9.
146. Wu, Z.L., et al., Altered behavior and long-term potentiation in type I adenylyl cyclase mutant mice. Proc Natl Acad Sci U S A, 1995. 92(1): p. 220-4.
147. Abel, T., et al., Genetic demonstration of a role for PKA in the late phase of LTP and in hippocampus-based long-term memory. Cell, 1997. 88(5): p. 615-26.
148. Arnsten, A.F., et al., Protein kinase A as a therapeutic target for memory disorders: rationale and challenges. Trends Mol Med, 2005. 11(3): p. 121-8.
149. Blokland, A., R. Schreiber, and J. Prickaerts, Improving memory: a role for phosphodiesterases. Curr Pharm Des, 2006. 12(20): p. 2511-23.
150. Boess, F.G., et al., Inhibition of phosphodiesterase 2 increases neuronal cGMP, synaptic plasticity and memory performance. Neuropharmacology, 2004. 47(7): p. 1081-92.
151. Prickaerts, J., et al., Phosphodiesterase type 5 inhibition improves early memory consolidation of object information. Neurochem Int, 2004. 45(6): p. 915-28.
152. van der Staay, F.J., et al., The novel selective PDE9 inhibitor BAY 73-6691 improves learning and memory in rodents. Neuropharmacology, 2008. 55(5): p. 908-18.
153. Cuadrado-Tejedor, M., et al., Sildenafil restores cognitive function without affecting beta-amyloid burden in a mouse model of Alzheimer's disease. Br J Pharmacol, 2011. 164(8): p. 2029-41.
154. Puzzo, D., et al., Phosphodiesterase 5 inhibition improves synaptic function, memory, and amyloid-beta load in an Alzheimer's disease mouse model. J Neurosci, 2009. 29(25): p. 8075-86.
126
155. Garcia-Barroso, C., et al., Tadalafil crosses the blood-brain barrier and reverses
cognitive dysfunction in a mouse model of AD. Neuropharmacology, 2013. 64: p. 114-23.
156. Yuste, R. and T. Bonhoeffer, Morphological changes in dendritic spines associated with long-term synaptic plasticity. Annu Rev Neurosci, 2001. 24: p. 1071-89.
157. Frey, U. and R.G. Morris, Synaptic tagging and long-term potentiation. Nature, 1997. 385(6616): p. 533-6.
158. Matsuzaki, M., et al., Structural basis of long-term potentiation in single dendritic spines. Nature, 2004. 429(6993): p. 761-6.
159. Yuste, R., A. Majewska, and K. Holthoff, From form to function: calcium compartmentalization in dendritic spines. Nat Neurosci, 2000. 3(7): p. 653-9.
160. Fu, M. and Y. Zuo, Experience-dependent structural plasticity in the cortex. Trends Neurosci, 2011. 34(4): p. 177-87.
161. Bailey, C.H. and M. Chen, Time course of structural changes at identified sensory neuron synapses during long-term sensitization in Aplysia. J Neurosci, 1989. 9(5): p. 1774-80.
162. Geinisman, Y., et al., Remodeling of hippocampal synapses after hippocampus-dependent associative learning. J Comp Neurol, 2000. 417(1): p. 49-59.
163. Penzes, P., et al., Dendritic spine pathology in neuropsychiatric disorders. Nat Neurosci, 2011. 14(3): p. 285-93.
164. Dorostkar, M.M., et al., Analyzing dendritic spine pathology in Alzheimer's disease: problems and opportunities. Acta Neuropathol, 2015. 130(1): p. 1-19.
165. Bailey, C.H., E.R. Kandel, and K.M. Harris, Structural Components of Synaptic Plasticity and Memory Consolidation. Cold Spring Harb Perspect Biol, 2015. 7(7): p. a021758.
166. Neves, G., S.F. Cooke, and T.V. Bliss, Synaptic plasticity, memory and the hippocampus: a neural network approach to causality. Nat Rev Neurosci, 2008. 9(1): p. 65-75.
167. Whitlock, J.R., et al., Learning induces long-term potentiation in the hippocampus. Science, 2006. 313(5790): p. 1093-7.
168. Pastalkova, E., et al., Storage of spatial information by the maintenance mechanism of LTP. Science, 2006. 313(5790): p. 1141-4.
169. Oh, W.C., T.C. Hill, and K. Zito, Synapse-specific and size-dependent mechanisms of spine structural plasticity accompanying synaptic weakening. Proc Natl Acad Sci U S A, 2013. 110(4): p. E305-12.
127
170. Okamoto, K., M. Bosch, and Y. Hayashi, The roles of CaMKII and F-actin in the
structural plasticity of dendritic spines: a potential molecular identity of a synaptic tag? Physiology (Bethesda), 2009. 24: p. 357-66.
171. Borovac, J., et al. Bidirectional role of postsynaptic cAMP and cGMP in synaptic plasticity and memory. in Neuroscience Meeting. 2017. Washington DC: Society for Neuroscience.
172. Enna, S.J. and E.W. Karbon, Receptor Regulation - Evidence for a Relationship between Phospholipid-Metabolism and Neurotransmitter Receptor-Mediated Camp Formation in Brain. Trends in pharmacological sciences, 1987. 8(1): p. 21-24.
173. Frey, U., Y.Y. Huang, and E.R. Kandel, Effects of Camp Simulate a Late-Stage of Ltp in Hippocampal Ca1 Neurons. Science, 1993. 260(5114): p. 1661-1664.
174. Greengard, P., et al., Enhancement of the Glutamate Response by Camp-Dependent Protein-Kinase in Hippocampal-Neurons. Science, 1991. 253(5024): p. 1135-1138.
175. Kandel, E.R., The molecular biology of memory: cAMP, PKA, CRE, CREB-1, CREB-2, and CPEB. Mol Brain, 2012. 5: p. 14.
176. Shelly, M., et al., Local and long-range reciprocal regulation of cAMP and cGMP in axon/dendrite formation. Science, 2010. 327(5965): p. 547-52.
177. Ryu, M.-H., et al., Natural and Engineered Photoactivated Nucleotidyl Cyclases for Optogenetic Applications. Journal of Biological Chemistry, 2010. 285(53): p. 41501-41508.
178. Helmchen, F. and W. Denk, Deep tissue two-photon microscopy. Nat Methods, 2005. 2(12): p. 932-40.
179. Fedoryak, O.D., et al., Synthesis of a caged glutamate for efficient one- and two-photon photorelease on living cells. Chem Commun (Camb), 2005(29): p. 3664-6.
180. Zito, K., et al., Rapid functional maturation of nascent dendritic spines. Neuron, 2009. 61(2): p. 247-58.
181. Stein, I.S. and K. Zito, Dendritic Spine Elimination: Molecular Mechanisms and Implications. Neuroscientist, 2018: p. 1073858418769644.
182. Noguchi, J., et al., State-dependent diffusion of actin-depolymerizing factor/cofilin underlies the enlargement and shrinkage of dendritic spines. Sci Rep, 2016. 6: p. 32897.
183. Seamon, K.B., W. Padgett, and J.W. Daly, Forskolin: unique diterpene activator of adenylate cyclase in membranes and in intact cells. Proc Natl Acad Sci U S A, 1981. 78(6): p. 3363-7.
128
184. Hartzell, H.C. and R. Fischmeister, Opposite effects of cyclic GMP and cyclic AMP
on Ca2+ current in single heart cells. Nature, 1986. 323(6085): p. 273-5.
185. Beavo, J.A., Cyclic nucleotide phosphodiesterases: functional implications of multiple isoforms. Physiol Rev, 1995. 75(4): p. 725-48.
186. Zaccolo, M. and M.A. Movsesian, cAMP and cGMP signaling cross-talk: role of phosphodiesterases and implications for cardiac pathophysiology. Circ Res, 2007. 100(11): p. 1569-78.
187. Horstrup, K., et al., Phosphorylation of focal adhesion vasodilator-stimulated phosphoprotein at Ser157 in intact human platelets correlates with fibrinogen receptor inhibition. Eur J Biochem, 1994. 225(1): p. 21-7.
188. Feil, R. and T. Kleppisch, NO/cGMP-dependent modulation of synaptic transmission. Handb Exp Pharmacol, 2008(184): p. 529-60.
189. Wu, J., et al., Evidence for involvement of the cGMP-protein kinase G signaling system in the induction of long-term depression, but not long-term potentiation, in the dentate gyrus in vitro. J Neurosci, 1998. 18(10): p. 3589-96.
190. Britt, J.P., R.A. McDevitt, and A. Bonci, Use of channelrhodopsin for activation of CNS neurons. Curr Protoc Neurosci, 2012. Chapter 2: p. Unit2 16.
191. Ryu, M.H., et al., Natural and engineered photoactivated nucleotidyl cyclases for optogenetic applications. J Biol Chem, 2010. 285(53): p. 41501-8.
192. Shaner, N.C., et al., Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat Biotechnol, 2004. 22(12): p. 1567-72.
193. Russwurm, M., et al., Design of fluorescence resonance energy transfer (FRET)-based cGMP indicators: a systematic approach. Biochem J, 2007. 407(1): p. 69-77.
194. Mulkey, R.M., et al., Involvement of a calcineurin/inhibitor-1 phosphatase cascade in hippocampal long-term depression. Nature, 1994. 369(6480): p. 486-8.
195. Huang, K.X. and H.K. Paudel, Ser67-phosphorylated inhibitor 1 is a potent protein phosphatase 1 inhibitor. Proc Natl Acad Sci U S A, 2000. 97(11): p. 5824-9.
196. Bollen, E., et al., Improved long-term memory via enhancing cGMP-PKG signaling requires cAMP-PKA signaling. Neuropsychopharmacology, 2014. 39(11): p. 2497-505.
197. Prickaerts, J., et al., cGMP, but not cAMP, in rat hippocampus is involved in early stages of object memory consolidation. Eur J Pharmacol, 2002. 436(1-2): p. 83-7.
198. Brandon, E.P., et al., Hippocampal long-term depression and depotentiation are defective in mice carrying a targeted disruption of the gene encoding the RI beta
129
subunit of cAMP-dependent protein kinase. Proc Natl Acad Sci U S A, 1995. 92(19): p. 8851-5.
199. Madronal, N., et al., Rapid erasure of hippocampal memory following inhibition of dentate gyrus granule cells. Nat Commun, 2016. 7: p. 10923.
200. Palmer, A. and M. Good, Hippocampal synaptic activity, pattern separation and episodic-like memory: implications for mouse models of Alzheimer's disease pathology. Biochem Soc Trans, 2011. 39(4): p. 902-9.
201. Jacobsen, J.S., et al., Early-onset behavioral and synaptic deficits in a mouse model of Alzheimer's disease. Proc Natl Acad Sci U S A, 2006. 103(13): p. 5161-6.
202. Weeden, C.S., et al., The role of the ventral dentate gyrus in anxiety-based behaviors. Neurobiol Learn Mem, 2015. 118: p. 143-9.
203. Kheirbek, M.A., et al., Differential control of learning and anxiety along the dorsoventral axis of the dentate gyrus. Neuron, 2013. 77(5): p. 955-68.
204. Zou, D., et al., DREADD in parvalbumin interneurons of the dentate gyrus modulates anxiety, social interaction and memory extinction. Curr Mol Med, 2016. 16(1): p. 91-102.
205. Antonova, I., et al., Rapid increase in clusters of presynaptic proteins at onset of long-lasting potentiation. Science, 2001. 294(5546): p. 1547-50.
206. Roy, D.S., et al., Memory retrieval by activating engram cells in mouse models of early Alzheimer's disease. Nature, 2016. 531(7595): p. 508-12.
207. Allen, T.A. and N.J. Fortin, The evolution of episodic memory. Proc Natl Acad Sci U S A, 2013. 110 Suppl 2: p. 10379-86.
208. Vorhees, C.V. and M.T. Williams, Assessing spatial learning and memory in rodents. ILAR J, 2014. 55(2): p. 310-32.
209. Feil, R., et al., cGMP-dependent protein kinase I, the circadian clock, sleep and learning. Commun Integr Biol, 2009. 2(4): p. 298-301.
210. Wincott, C.M., et al., cGMP-dependent protein kinase type II knockout mice exhibit working memory impairments, decreased repetitive behavior, and increased anxiety-like traits. Neurobiol Learn Mem, 2014. 114: p. 32-9.
211. Wincott, C.M., et al., Spatial memory deficits and motor coordination facilitation in cGMP-dependent protein kinase type II-deficient mice. Neurobiol Learn Mem, 2013. 99: p. 32-7.
212. Werner, C., et al., Importance of NO/cGMP signalling via cGMP-dependent protein kinase II for controlling emotionality and neurobehavioural effects of alcohol. Eur J Neurosci, 2004. 20(12): p. 3498-506.
130
213. Rutten, K., et al., The selective PDE5 inhibitor, sildenafil, improves object memory in
Swiss mice and increases cGMP levels in hippocampal slices. Behav Brain Res, 2005. 164(1): p. 11-6.
214. Kurt, M., et al., Effect of sildenafil on anxiety in the plus-maze test in mice. Pol J Pharmacol, 2004. 56(3): p. 353-7.
215. Lueptow, L.M., C.G. Zhan, and J.M. O'Donnell, Cyclic GMP-mediated memory enhancement in the object recognition test by inhibitors of phosphodiesterase-2 in mice. Psychopharmacology (Berl), 2016. 233(3): p. 447-56.
216. Lu, Y.F., E.R. Kandel, and R.D. Hawkins, Nitric oxide signaling contributes to late-phase LTP and CREB phosphorylation in the hippocampus. J Neurosci, 1999. 19(23): p. 10250-61.
217. Paul, C., et al., cGMP-dependent protein kinase type I promotes CREB/CRE-mediated gene expression in neurons of the lateral amygdala. Neurosci Lett, 2010. 473(2): p. 82-6.
218. Monfort, P., et al., Long-term potentiation in hippocampus involves sequential activation of soluble guanylate cyclase, cGMP-dependent protein kinase, and cGMP-degrading phosphodiesterase. J Neurosci, 2002. 22(23): p. 10116-22.
219. Li, H., et al., Experience-dependent modification of a central amygdala fear circuit. Nat Neurosci, 2013. 16(3): p. 332-9.
220. Zacharias, D.A., et al., Partitioning of lipid-modified monomeric GFPs into membrane microdomains of live cells. Science, 2002. 296(5569): p. 913-6.
221. Niwa, H., K. Yamamura, and J. Miyazaki, Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene, 1991. 108(2): p. 193-9.
222. Stierl, M., et al., Light modulation of cellular cAMP by a small bacterial photoactivated adenylyl cyclase, bPAC, of the soil bacterium Beggiatoa. J Biol Chem, 2011. 286(2): p. 1181-8.
223. Honda, A., et al., Spatiotemporal dynamics of guanosine 3',5'-cyclic monophosphate revealed by a genetically encoded, fluorescent indicator. Proc Natl Acad Sci U S A, 2001. 98(5): p. 2437-42.
224. Hille, C., et al., Two-photon fluorescence lifetime imaging of intracellular chloride in cockroach salivary glands. Photochem Photobiol Sci, 2009. 8(3): p. 319-27.
225. Yasuda, R., et al., Supersensitive Ras activation in dendrites and spines revealed by two-photon fluorescence lifetime imaging. Nat Neurosci, 2006. 9(2): p. 283-91.
226. Puzzo, D., et al., Amyloid-beta peptide inhibits activation of the nitric oxide/cGMP/cAMP-responsive element-binding protein pathway during hippocampal synaptic plasticity. J Neurosci, 2005. 25(29): p. 6887-97.
131
227. Shoji, H., et al., Age-related changes in behavior in C57BL/6J mice from young
adulthood to middle age. Mol Brain, 2016. 9: p. 11.
228. Ageta-Ishihara, N., et al., Chronic overload of SEPT4, a parkin substrate that aggregates in Parkinson's disease, causes behavioral alterations but not neurodegeneration in mice. Mol Brain, 2013. 6: p. 35.