Download Ionotropic Glutamate Receptors in Aplysia californica and Molecular

University of Miami
Scholarly Repository
Open Access Dissertations
Electronic Theses and Dissertations
2017-04-25
Ionotropic Glutamate Receptors in Aplysia
californica and Molecular Changes in the Aging
Nervous System
Justin B. Greer
University of Miami, [email protected]
Follow this and additional works at: http://scholarlyrepository.miami.edu/oa_dissertations
Recommended Citation
Greer, Justin B., "Ionotropic Glutamate Receptors in Aplysia californica and Molecular Changes in the Aging Nervous System" (2017).
Open Access Dissertations. 1819.
http://scholarlyrepository.miami.edu/oa_dissertations/1819
This Open access is brought to you for free and open access by the Electronic Theses and Dissertations at Scholarly Repository. It has been accepted for
inclusion in Open Access Dissertations by an authorized administrator of Scholarly Repository. For more information, please contact
[email protected].
UNIVERSITY OF MIAMI
IONOTROPIC GLUTAMATE RECEPTORS IN APLYSIA CALIFORNICA AND
MOLECULAR CHANGES IN THE AGING NERVOUS SYSTEM
By
Justin B. Greer
A DISSERTATION
Submitted to the Faculty
of the University of Miami
in partial fulfillment of the requirements for
the degree of Doctor of Philosophy
Coral Gables, Florida
May 2017
©2017
Justin B. Greer
All Rights Reserved
UNIVERSITY OF MIAMI
A dissertation submitted in partial fulfillment of
the requirements for the degree of
Doctor of Philosophy
IONOTROPIC GLUTAMATE RECEPTORS IN APLYSIA
CALIFORNICA AND MOLECULAR CHANGES IN THE
AGING NERVOUS SYSTEM
Justin B. Greer
Approved:
________________
Lynne Fieber, Ph.D.
Associate Professor of Marine
Biology and Ecology
_________________
Michael Schmale, Ph.D.
Professor of Marine
Biology and Ecology
________________
Marjorie Oleksiak, Ph.D.
Associate Professor of Marine
Biology and Ecology
_________________
Sathya Puthanveettil, Ph.D.
Professor of Neuroscience
Scripps Research Institute
________________
Sawsan Khuri, Ph.D.
Research Assistant Professor
Department of Computer Science
_________________
Guillermo Prado, Ph.D.
Dean of the Graduate School
GREER, JUSTIN B.
(Ph.D., Marine Biology and Ecology)
Ionotropic Glutamate Receptors in Aplysia californica
(May 2017)
and Molecular Changes in the Aging Nervous System
Abstract of a dissertation at the University of Miami.
Dissertation supervised by Professor Lynne Fieber.
No. of pages in text. (157)
Aplysia californica is a marine snail with well-defined neural circuits and a relatively
short life span of one year, making it a useful model for studies of neural function and
aging. Many of these neural circuits likely use L-Glutamate as the excitatory
neurotransmitter. The goal of this dissertation was to molecularly characterize ionotropic
L-Glutamate receptors in A. californica, and describe molecular changes in these
receptors during aging in well-defined neural circuits.
The evolutionary relationship between ionotropic L-Glutamate receptors of A.
californica and vertebrates was first studied to evaluate the relevance of A. californica as
a model of ionotropic L-Glutamate receptor function. Genes belonging to the N-methylD-aspartate receptor (NMDAR) and α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic
acid receptor (AMPAR) subtypes were found to be particularly well-conserved between
A. californica and vertebrates. Both subtypes are critical for synaptic plasticity
underlying learning and memory. These results suggest conserved function of NMDAR
and AMPAR over large evolutionary distances, thus making A. californica a suitable
model for studying their function. Expression of all identified ionotropic L-Glutamate
receptors in all A. californica nervous system tissues emphasized their important role in
the nervous system.
Next, age-related changes in L-Glutamate receptor expression were examined in
two sensory neuron clusters. Several NMDAR subtype genes exhibited reduced
expression in aged animals compared to sexually mature animals. This suggested that
reduced NMDAR in sensory neurons may contribute to previously described declines in
L-Glutamate current amplitude and reduced reflexes of aged animals.
Finally, whole transcriptome expression changes in aged sensory neurons were
studied using RNASeq. Transcription of ion channel genes and genes related to nervous
system function were significantly affected in aged compared to mature animals. These
alterations may negatively affect synaptic transmission and contribute to previously
observed deficits in nervous system function of aged A. californica.
TABLE OF CONTENTS
Page
LIST OF FIGURES .....................................................................................................
v
LIST OF TABLES ...................................................................................................... viii
Chapter
1
INTRODUCTION ...........................................................................................
1.1 Ionotropic glutamate receptors in the nervous system ...............................
1.2 Aplysia as a model for aging and learning and memory ...........................
1
1
6
2
PHYLOGENETIC ANALYSIS OF IONOTROPIC GLUTAMATE
RECEPTOR GENES IN THE BILATERIA, WITH SPECIAL NOTES ON
APLYSIA CALIFORNICA ............................................................................
2.1 Summary ....................................................................................................
2.2 Background ................................................................................................
2.3 Materials and Methods ...............................................................................
2.4 Results ........................................................................................................
2.5 Discussion ..................................................................................................
15
15
16
19
22
31
ALTERED EXPRESSION OF IONOTROPIC L-GLUTAMATE
RECEPTORS IN AGED SENSORY NEURONS OF APLYSIA ..................
3.1 Summary ....................................................................................................
3.2 Background ................................................................................................
3.3 Materials and Methods ...............................................................................
3.4 Results ........................................................................................................
3.5 Discussion ..................................................................................................
47
47
47
52
57
63
3
4
TRANSCRIPTIONAL CHANGES DURING AGING IN SENSORY
NEURONS INNERVATING TWR IN APLYSIA ........................................ 81
4.1 Summary .................................................................................................... 81
4.2 Background ................................................................................................ 81
4.3 Materials and Methods ............................................................................... 88
4.4 Results ........................................................................................................ 95
4.5 Discussion .................................................................................................. 100
5
CONCLUSIONS ............................................................................................. 123
REFERENCES ............................................................................................................ 132
iii
Appendix
A CHAPTER 2 SUPPLEMENTARY FIGURES ............................................... 151
B CHAPTER 4 SUPPLEMENTARY DATA FILES ......................................... 157
iv
LIST OF FIGURES
Page
Chapter 2
2.1 Phylogeny of species used for iGluR analysis ...........................................
42
2.2 Bootstrap consensus phylogenetic tree ......................................................
43
2.3 Conserved motifs in Aplysia and H. sapiens .............................................
44
2.4 Hydrophobicity plot of Grin1 in H. sapiens and Aplysia...........................
45
2.5 iGluR expression levels in Aplysia ganglia ...............................................
46
Chapter 3
3.1 Behavioral and weight correlates of aging in two cohorts of Aplysia .......
71
3.2 Differential expression of iGluR subunits within abdominal ganglia and
sensory neuronal clusters .................................................................................
72
3.3 Calculated number of iGluR transcripts in different ganglia .....................
73
3.4 Relative expression of iGluR subunits in PVC sensory neurons of Mature and
Aged II animals from two cohorts ...................................................................
74
3.5 Absolute transcript quantification of iGluR subunits in PVC sensory neurons
of Mature and Aged II animals from cohort 1 .................................................
75
3.6 Relative expression of iGluR subunits in BSC sensory neurons of Mature and
Aged II animals from two cohorts ...................................................................
76
3.7 Absolute transcript quantification of iGluR subunits in BSC sensory neurons
of Mature and Aged II animals from cohort 1 .................................................
77
3.8 Calculated number of iGluR transcripts in Mature and Aged II
whole abdominal ganglion of cohort 1 ............................................................
v
78
3.9 Percent composition of NMDAR subtype subunits ...................................
79
3.10 Percent composition of AMPAR subtype subunits .................................
80
Chapter 4
4.1 Behavioral assessments of aging in cohorts used for RNASeq ................. 112
4.2 Principal component analysis (PCA) of the first two principal components in
M and AII PVC neurons .................................................................................. 113
4.3 Overlap of DE genes identified with STAR and Tophat2 ......................... 114
4.4 Scatterplot of mean expression of genes analyzed by STAR vs. their
log-fold change ................................................................................................ 115
4.5 Heatmap of the top 50 most significantly DE genes from PVC neuron clusters
of M and AII animals ....................................................................................... 116
4.6 Expression of iGluR subunits in M and AII PVC neurons ........................ 117
4.7 Composition of NMDAR and AMPAR in M and AII PVC neurons ........ 118
4.8 The pathways and activities most affected by aging ................................. 119
4.9 Mean expression vs log2 fold change for ion channel associated genes ...
120
4.10 Mean expression vs log2 fold change for stress response genes .............. 121
4.11 Comparison of DE genes in RNASeq and qPCR analyses ...................... 122
Appendix A
S2.1 NMDAR subtype only tree ...................................................................... 151
S2.2 AMPAR subtype only tree....................................................................... 152
S2.3 Kainate receptor subtype only tree .......................................................... 153
S2.4 Aplysia iGluR protein tree ....................................................................... 154
vi
S2.5 Representative AMPA subunit TMD Hydrophobicity ............................ 155
S2.6 Representative kainate subunit TMD Hydrophobicity ............................ 156
vii
LIST OF TABLES
Page
Chapter 2
2.1 Primers used for qPCR...............................................................................
38
2.2 Placement of Protostomia iGluR into subtypes based on phylogenies ......
39
2.3 Number of iGluR genes in the protostomes and vertebrates .....................
40
2.4 Sequence similarities in transmembrane domains of Grin1 between different
protostomes and H. sapiens .............................................................................
41
Chapter 4
4.1 Primer sequences for qPCR ....................................................................... 110
4.2 Read depth and mapping statistics using STAR ........................................ 111
viii
Chapter 1:
Introduction
1.1 Ionotropic glutamate receptors in the nervous system
The mechanisms underlying learning and memory are vastly complex and involve
a multitude of molecular changes that may result in learned behaviors. Mammalian
systems may be difficult to work use to understand learned behaviors due to highly
complex nervous systems; the human brain consists of one trillion neurons and highly
complicated neuronal networks, making it difficult to elucidate learning and memory
processes. The mammalian hippocampus is believed to be heavily involved in learning,
but contains ~40 million neurons (Gundersen, et al. 2001). Tracing how sensory
information is received by the hippocampus and how this information is transmitted into
behavior through this large neural network is immensely difficult, even without the added
complications of learning and memory events. These confounding factors facilitated the
need for a simpler model system to understand learning and memory basics. This led Eric
Kandel in the early 1960’s to study Aplysia californica (Aplysia), culminating in the 2000
Nobel Prize in Physiology and Medicine for his work on the physiological basis of
memory formation and storage.
Aplysia is a model system for learning and memory with several distinct
advantages compared to mammals. The Aplysia nervous system contains ~20,000
neurons, and simple forms of behavior can be followed from stimulus to behavior with
less complexity and involvement of interneurons. Aplysia neurons are some of the largest
in the animal kingdom at up to of 1 mm in diameter, can be seen with the naked eye, and
are easily reidentifiable in consecutive individual animals, allowing for direct
1
2
electrophysiological recordings from individual cells. Large neurons also allow for
nucleic acid isolation from few cells, allowing for targeted studies on subsets of neurons.
Glutamate (L-Glu) is the principal fast excitatory neurotransmitter in the central
nervous system (CNS) of nearly all animals. In the Aplysia CNS, there is abundant
evidence that these sensory neurons use L-Glu as their primary excitatory
neurotransmitter. The nerve fibers that connect sensory neurons to other neurons are
strongly immunoreactive for L-Glu (Levenson, Sherry, et al. 2000). These sensory
neurons have high expression of L-Glu-like receptors (Dale and Kandel 1993; Ha, et al.
2006), and their [L-Glu] is high at 29 mM, compared to 3 mM in unidentified neurons of
the abdominal ganglion (Drake, et al. 2005), suggesting that L-Glu mediated responses
are prevalent in the Aplysia nervous system.
L-Glu receptors are classified into two types, ionotropic and metabotropic
receptors. Metabotropic L-Glu receptors (mGluR) are G-protein-coupled receptors in
which binding of L-Glu activates intracellular cascades and modification of intracellular
proteins; they are not ion channels. In contrast, ionotropic L-Glu receptors (iGluR) are
ion channels that convey the majority of fast excitatory signal transmission modulated by
L-Glu and have been implicated in most aspects of CNS development and function
(Traynelis, et al. 2010a). The binding of L-Glu to iGluR opens transmembrane ion
channels that allow ion flux across the plasma membrane and depolarization of the
postsynaptic cell. Sufficient agonist-receptor binding results in superthreshold
depolarization that triggers an action potential, thereby transmitting information. iGluR
also play a significant role in synaptic plasticity associated with learning (Roche, et al.
1994; Kandel 2001; Paoletti, et al. 2013).
3
iGluR are divided into three subtypes; N-methyl-D-aspartate (NMDA), kainate,
and α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), based on
selective agonists. The latter two subtypes are known as non-NMDA receptors (nonNMDAR). Overall amino acid identity of iGluR subunits across the three families in
mammals is 20-30%, but each contain common structural features that place them
together into a single large superfamily (Kew and Kemp 2005). Function and expression
varies between subtypes, but there are several common features of all iGluR. In
eukaryotes iGluR are tetrameric in structure, consisting of four subunits, and are formed
as a dimer of dimers (Tichelaar, et al. 2004). Each complete receptor contains an agonistbinding domain, three transmembrane domains, an extracellular N terminus, and an
intracellular C terminus (Traynelis, et al. 2010a). Studies on receptor structure have been
limited primarily to vertebrates, and it is unknown if Aplysia iGluR are comprised of this
same structure. Insights into composition of these receptors are an important step in
furthering the Aplysia model.
Although they are related in structure, there are several distinct
electrophysiological and activation requirements that are unique to individual iGluR
subtypes. Activation kinetics of non-NMDAR show that activation, deactivation, and
desensitization occur on a millisecond timescale, while NMDAR display these
characteristics much slower, on a timescale of tens to hundreds of milliseconds (Hansen,
et al. 2007; Paoletti and Neyton 2007). AMPAR are permeable to Na+, K+, as well as
infrequently permeable to Ca2+, depending on the subunit structure, with most AMPAR
containing the Ca2+ impermeable Gria2 subunits, and are thus impermeable to Ca2+
4
(Greger, et al. 2007). Conversely, Ca2+ is the principal ion transported by NMDAR in
vertebrate models (Mayer and Westbrook 1987).
In mammals, NMDAR are obligatory heterotetramers composed of various
arrangements of three subunits, Grin1, Grin2, and Grin3 (Glutamate Receptor
Ionotropic/NMDA). This may be due to the fact that, unlike AMPAR, NMDAR require
binding of the co-agonist glycine for activation. In mammals, complete NMDAR are
comprised of two Grin1 subunits responsible for glycine binding (or d-serine), and either
two Grin2 or Grin3 subunits binding L-Glu, although NMDAR containing all three
subunits have also been reported (Traynelis, et al. 2010a). AMPAR preferentially form
heterotetramers, but are also functional as a homotetramer, containing four of the same
subunit (Sobolevsky, et al. 2009).
Vertebrate NMDAR are also subject to allosteric modulation by extracellular
Mg2+, which blocks the channel constitutively in a voltage-dependent manner (Mayer, et
al. 1984). The cation must be excluded by depolarization for Ca2+ to enter the channel.
They are also modulated by other small molecules or ions, such as H+ and Zn2+, which
are endogenously present in the CNS and act as allosteric modulators (Paoletti 2011).
Similar to the work on receptor structure, the majority of these findings were
demonstrated in vertebrates, and modifications, if they occur, are unclear in Aplysia.
High Mg2+ concentrations in seawater make Mg2+ unlikely to be an effective mechanism
for constitutive channel block without very high energy cost to remove it. Seawater
contains large amounts of glycine that, similarly, make it unlikely to be an effective coagonist. Aplysia are osmoconformers not believed to selectively exclude ions, resulting in
high endogenous Mg2+ and glycine levels in their CNS.
5
The Grin1 subunit is encoded by a single gene in mammals, and contains a
varying number of splices depending on species, with as many as eight splice variants in
humans and two in mice (Paoletti 2011). Grin2 subunits are represented by four
members, each encoded by its own gene, and Grin3 has two members each encoded by a
separate gene. Currently, two Grin-like receptors and one Grin2-like receptor have been
cloned in Aplysia (Ha, et al. 2006). Analysis in Prosite (Letunic, et al. 2012) predicts
binding sites for both L-Glu and glycine in the Aplysia Grin1-like genes. Grin3-like
subunits have yet to be found in Aplysia.
Incorporation of various subunit compositions plays a major role in receptor
functional properties, and changes in expression of subunit (and isoform) mRNA is
prevalent during development (Akazawa, et al. 1994; Laurie and Seeburg 1994).
Functionally, permeability of NMDAR to Ca2+ as well as the extent of Mg2+ block is
directly related to the subunits that comprise the receptor (Schneggenburger 1996). The
intracellular N-terminal domain, ~400 amino acids, is thought to be the major
determinant of subtype-specific assembly (Paoletti 2011).
For several reasons, classification of iGluR subunits into subtypes is less clear in
invertebrates. First, there is high divergence of these sequences from their vertebrate
counterparts, making molecular phylogenetics difficult because invertebrate iGluR
cluster together as a separate node from vertebrate iGluR using full-length sequences.
Further confounding classification of Aplysia iGluR, subtype specific agonists are less
effective in Aplysia neurons than in vertebrates. The structure of iGluR in Aplysia are
sufficiently different that neurons do not respond to AMPA as an agonist (Trudeau and
Castellucci 1993; Carlson and Fieber 2012). Kainate analogs do elicit depolarization and
6
non-desensitizing inward current in glutamatergic BSC neurons and abdominal ganglion
neurons (Trudeau and Castellucci 1993; Wang, et al. 2013), and NMDA elicits small
amplitude currents in BSC neurons (Carlson and Fieber 2012).
Of the three iGluR subtypes, AMPAR and NMDAR have been implicated in
underlying physiological changes that accompany learning and memory. Aplysia and
other invertebrate model organisms have been integral in furthering research on the
underlying mechanisms responsible for learning and memory. However, there is a current
lack of evidence for the subunits that may play the roles of AMPAR and NMDAR in
invertebrates. Identification of these subunits could allow for could allow for targeted
studies of these iGluR subunits to gain further insights into iGluR physiology in
invertebrate nervous systems.
1.2 Aplysia as a model for aging and learning and memory
Aging is associated with sensory and motoneuron function impairments that result
in deterioration of cognitive function and reflex systems in aged animals. Significant
neuron loss is not proposed to be a major component of reduced neuronal function with
age (Hof and Morrison 2004). Rather, a reduced number of synaptic contacts and reduced
dendritic spine density may alter the strength of synaptic connections (Wallace, et al.
2007; Dumitriu, et al. 2010). Reduced postsynaptic density in axiospinous synapses in
hippocampal neurons has been implicated in learning impairments in aged rats
(Nicholson, et al. 2004). Reduced excitatory post synaptic potential amplitude in the aged
hippocampus suggests that neuronal communication is reduced during aging (Landfield,
et al. 1986; Boxer, et al. 1988; Deupree, et al. 1993). Significant decreases in conduction
7
velocity in several areas of the brain may also be caused by changes in the myelin sheath
(Kanda, et al. 1986; Boxer, et al. 1988; Peters 2002). Taken together, these changes in
structure and function of neuronal pathways may lead to age-related impairments in the
aging brain.
Aplysia provides a unique model system with several advantages that make it an
excellent model for studies of aging at the behavioral and physiological levels. At the
National Resource for Aplysia animals have a lifespan of approximately one year,
allowing for study of both mature and aged animals in a relatively short period of time
(Gerdes and Fieber 2006). Sexual maturity occurs at ~7-8 months at 13-15°C at a density
of 5 animals per cage, but this can be manipulated with changes in either temperature or
cage density (Stommes, et al. 2005). Body length and weight increase linearly with age
through first sexual maturity, then plateau as food mass in and spawn mass out become
approximately equal. Egg mass production reaches its peak ~1-2 months after first sexual
maturity (Capo, et al. 2003; Gerdes and Fieber 2006). A decrease in body mass and
reduced egg production in the months preceding death signify the onset of senescence
(Hirsch and Peretz 1984; Gerdes and Fieber 2006). Arousal that results in increased heart
rate as a response to food stimuli was reduced in aged Aplysia (Bailey, et al. 1983).
Given the simple nervous system of Aplysia, the underlying neuronal circuits of
several behaviors have been studied in detail (Pinsker, et al. 1970; Byrne, Castellucci and
Kandel 1978; Peretz, et al. 1982; Walters, et al. 1983a, b; Peretz, et al. 1984). Many of
these behaviors are altered in aged animals, and physiological changes in the
corresponding neural circuits have been investigated. In the gill-withdrawal reflex
(GWR), a weak stimulus applied to the siphon causes the siphon and gill to withdraw into
8
the mantle for protection. GWR responsiveness to a stimulus is reduced in aged animals
and habituation is impaired compared to young animals (Rattan and Peretz 1981).
Physiologically, retraction of the siphon into the mantle is innervated by 24 sensory
neurons that connect with 6 motoneurons in the abdominal ganglion that withdrawal the
siphon (Carew, et al. 1971; Byrne, Castellucci, Carew, et al. 1978). Impairment of GWR
in aged animals has been correlated with decreased input resistance and decreased size of
post synaptic potentials evoked by gill stimulation (Rattan and Peretz 1981). This
resulted in reduced excitability and synaptic communication in the GWR circuit and
provided evidence that reduced physiology in aged neurons directly involved in
behavioral deficits.
Aging of the tail withdrawal reflex (TWR) has also been investigated both
behaviorally and physiologically. TWR is a monosynaptic reflex in which an electrical or
mechanical stimulus to the tail results in retraction of the tail into the body as a defense
mechanism (Walters, et al. 1983a). The time to relax the tail after tail touch, a measure of
TWR, increased significantly in aged animals and amplitude of tail retraction decreased
(Kempsell and Fieber 2014).
Tail mechanosensory neurons located in the ventral caudal (PVC) region of the
pleural ganglion are activated in response to tail touch. PVC neurons innervate
monosynaptic connections to tail motoneurons in the ipsilateral pedal ganglion,
withdrawing the tail and completing the reflex (Walters, et al. 1983a). The underlying
physiology of decreased TWR with age indicated that aging of PVC sensory neurons may
be responsible, and may involve reduced responsiveness to glutamatergic agonists
(Kempsell and Fieber 2014). PVC neurons had robust, excitatory iGluR responses to
9
applied L-Glu and D-Asp that declined significantly during aging, whether studied in
isolated (cultured) neurons or in in situ preparations (Fieber, et al. 2010; Kempsell and
Fieber 2015b).
There are also alterations in behaviors and physiology of the biting reflex with
age. Biting amplitude was significantly reduced and latency of the response increased in
aged animals compared to animals at maturity. Buccal S cluster (BSC) sensory neurons
of the biting reflex in aged animals had reduced amplitude and current densities to L-Glu.
Thus, reduced current density and amplitude occured sensory neurons of TWR and the
biting reflex in aged animals in response to L-Glu. This suggests molecular changes in
iGluR that mediate glutamatergic responses. Decreased iGluR to bind L-Glu, a change in
sensitivity to agonist, or a change in receptor structure could lead to reduced glutamate
physiology in aged PVC and BSC sensory neurons.
In Aplysia, both the GWR and the TWR exhibit sensitization and habituation, two
forms of non-associative learning in which repetition of a stimulus results in changes in
the strength of the response. Sensitization results in amplification of the response, while
progressive diminution of a response constitutes habituation. These modifications, known
as synaptic plasticity, are the primary foundation of learning and memory. Sensitization
and habituation of the aforementioned neuronal circuits can be observed both
behaviorally and electrophysiologically in Aplysia. Behaviorally, repeated noxious
electrical stimulation of the tail results in an incrementally increased amplitude of TWR
that is characteristic of sensitization (Kempsell and Fieber 2014). Electrophysiologically,
sensitization leads to an increase in the number of spikes in the motor neuron, and an
increase in the amplitude of the monosynaptic excitatory postsynaptic potential (EPSP) in
10
the motoneuron (Carew, et al. 1981; Walters, et al. 1983b; Kempsell and Fieber 2015b).
The duration of the memory is a function of the number of repetitions of the stimulus and
occurs in two distinct forms. One or a few repetitions induces short-term facilitation
(STF) lasting minutes or hours, and many repetitions induce long-term facilitation (LTF)
that persists for 24 hours or several days. Each form of memory requires different
pathways for activation.
STF behavioral and electrophysiological changes are expressed in the presence of
protein synthesis inhibitors on both the pre- and postsynaptic neurons, suggesting fastacting mechanisms that are not dependent upon protein synthesis characterize STF
(Schwartz, et al. 1971). Electrical stimulation of presynaptic neurons resulted in a
doubling of cAMP in the neurons, and presynaptic STF was induced by injection of
cAMP without prior stimulation (Cedar and Schwartz 1972; Brunelli, et al. 1976).
Increased cAMP resulted in activation of the catalytic subunit of cAMP-dependent
protein kinase A (PKA). PKA phosphorylates presynaptic K+ channels, reducing the
number of active channels, resulting in greater Ca2+ influx per action potential, and
subsequent increases in amount of transmitter release by terminals of the sensory cell
(Castellucci, et al. 1980; Klein and Kandel 1980). STF memory appears to be primarily
due to these presynaptic mechanisms, resulting in enhanced synaptic strength through
increased neurotransmitter release at the synapse.
Application of the monoamine neurotransmitter serotonin (5-HT) mimics
behavioral training and resulted in increased cAMP and STF or LTF, depending on the
number of spaced applications (Cedar and Schwartz 1972). Later, identification of
serotonergic neurons in the Aplysia cerebral ganglia that participate in presynaptic
11
facilitation of sensory cells in the abdominal ganglia of the gill-withdrawal reflex
confirmed its role in vivo as the neurotransmitter modulating facilitation (Mackey, et al.
1989).
In contrast to STF, where enhancement of transmission is dependent only on
second messenger-induced phosphorylation, LTF is dependent on both phosphorylation
and protein synthesis. Inhibition of protein synthesis in the presynaptic or postsynaptic
neuron blocks induction of LTF, suggesting a different mechanism of action and
requirement for protein synthesis in both neurons (Castellucci, et al. 1989; Villareal, et al.
2007). 5-HT release from interneurons results in intracellular increases in both cAMP and
PKA, as in STF. With repeated pulses of 5-HT invoking LTF, however, PKA and
mitogen-activated protein kinase (MAPK) translocate to the nucleus and also
phosphorylate cAMP response element binding protein (CREB1) there (Bacskai, et al.
1993).
CREB-1 is a transcription factor and binds to a cAMP response element in the
promoter of target genes (Kandel 2001). CREB2, a repressor of CREB1-mediated gene
expression, acts jointly with CREB1 in coordinated regulation for initiation of LTF
(Bailey, et al. 1996). Two other positive transcriptional regulators, the CAAT box
enhancer-binding protein (C/EBP) and activation factor (Ap/AF), known as the
immediate response genes, are also activated with multiple applications of 5-HT
(Alberini, et al. 1994). Activation of immediate response genes stimulates transcription of
downstream genes that contribute to the formation of new synaptic connections between
sensory and motor neurons, a process not seen in STF (Bailey and Chen 1988b, a).
Sensorin, a sensory neuron-specific neuropeptide, is rapidly released from presynaptic
12
terminals, and expression is enhanced with repeated 5-HT applications (Hu, et al. 2004).
Sensorin, CREB1, and CREB2 show dynamic regulation during the first 24 hour after
LTF induction, and timing of expression appears to vary depending on the gene (Hart, et
al. 2011; Liu, et al. 2011).
Presynaptic structural changes observed in LTF in culture do not occur without
the presence of the motoneuron, suggesting signals from the postsynaptic neuron are a
requirement for expression of LTF (Glanzman, et al. 1990). Injection of 1,2-Bis-(2aminophenoxy)ethane-N,N,N’,N’-tetraacetic acid (BAPTA), a Ca2+ chelator, prevented
increased sensorin expression in the presynaptic cell body. This maneuver blocked LTF,
indicating elevated postsynaptic intracellular Ca2+ also is required for LTF (Cai, et al.
2008).
Regulation of postsynaptic NMDAR and AMPAR have been proposed to
significantly contribute to LTF. In Drosophila knockout of the Grin1 gene disrupted
olfactory learning, and the phenotype was rescued by transgene induction of Grin1 (Xia,
et al. 2005). This behavioral experiment was not verified electrophysiologically due to
the small size of Drosophila neurons. In the mammalian hippocampus, LTF in the
Schaffer collateral pathway is NMDA receptor dependent (Collingridge, et al. 1988). An
intracellular rise in Ca2+ is primarily mediated by NMDAR, and this rise in postsynaptic
intracellular Ca2+ is a necessary trigger for LTF (Malenka and Bear 2004).
Calcium/calmodulin-dependent protein kinase II (CaMKII) is required as a mediator for
NMDAR-dependent long term potentiation (Nicoll and Malenka 1999), an
electrophysiological proxy for the synaptic strengthening that is a component of learning.
Many studies have also found an increase in the number of functional AMPAR in the
13
synapse accompanying LTF (Trudeau and Castellucci 1995; Zhao, et al. 2003). Previous
work in Aplysia has shown that application of the AMPAR antagonist DNQX reversed
LTF that had been induced by 5-HT, indicating increased functional expression of
AMPAR in the motoneuron were necessary to maintain LTF (Chitwood, et al. 2001).
Induction of both STF and LTF are impaired in aged Aplysia. In the gill
withdrawal reflex long-term retention of habituation is impaired, and acquisition of LTF
of this reflex is absent in aged animals (Rattan and Peretz 1981). In TWR, sensitizing tail
shocks increased excitability of PVC neurons and enhanced neurotransmitter release at
the sensory to motoneuron synapse, strengthening the neuronal response to future stimuli.
In aged animals tail shocks failed to enhance TWR amplitude or duration (Kempsell and
Fieber 2015b). Additionally, enhanced excitation of PVC neurons and increased EPSP
amplitude in motoneurons is not observed. These data suggest regulation of genes vital
for induction of facilitation may be disrupted in the aged Aplysia nervous system.
Over the past 40 years the Aplysia model has been a pivotal tool in advancing our
understanding of the behavioral changes in the aging nervous system and their
corresponding electrophysiological alterations. This dissertation attempts to address
several outstanding questions that still remained. In Chapter 2, the evolutionary
relationship between vertebrate and invertebrate iGluR subunits is addressed. These data
give insights into the potential composition of iGluR in Aplysia, an important step in
understanding their similarity to well-studied vertebrate iGluR. Identification of subunits
most closely related to vertebrate iGluR subtypes allowed for hypotheses of iGluR
subunits that may be involved in learning and memory in Aplysia and other invertebrate
model organisms. In Chapter 3, transcriptional changes of iGluR subunits was examined
14
to identify subunits that may be involved in altered L-Glu physiology of aged sensory
neurons. In Chapter 4, whole-transcriptome profiling of PVC sensory neurons from
mature and aged Aplysia was assessed via RNASeq to characterize large-scale
transcriptional correlates of aging.
Together, these investigations provided a greater understanding iGluR in Aplysia
and the molecular effects of aging in the nervous system. This dissertation revealed that
reduced physiological function with age in sensory neurons may be the result of
significant declines in expression of components critical for neuronal homeostasis.
Previously described declines in L-Glu physiology of aged Aplysia were correlated with
reduced expression of iGluR that mediate the response. Significant alterations in ion
channel associated genes in aged animals may be crucial to observed sensory neuron
declines in excitability. NMDAR and AMPAR subtypes were found to be highly
conserved between Aplysia and vertebrates, providing evidence for conserved iGluR
function, thus making Aplysia a relevant model for iGluR function in the nervous system.
Chapter 2:
Phylogenetic analysis of ionotropic L-glutamate receptor genes in the Bilateria, with
special notes on Aplysia californica
2.1 Summary
Invertebrate organisms such as Aplysia californica (Aplysia) are well-studied
models for ionotropic L-Glutamate receptor (iGluR)-mediated function, yet no studies to
date have analyzed the evolutionary relationships between iGluR genes in Aplysia and
those in vertebrates, to identify genes that may mediate plasticity. We conducted a
thorough phylogenetic analysis spanning Bilateria to elucidate these relationships. The
expression status of iGluR genes in the Aplysia nervous system was also examined. Our
analysis shows that N-methyl-D-aspartate receptor (NMDAR) and α-Amino-3-hydroxy5-methyl-4-isoxazolepropionic acid receptor (AMPAR) subtypes were present in the
common bilaterian ancestor. NMDAR genes show very high conservation in motifs
responsible for forming the conductance pore of the ion channel. The number of
NMDAR subunits is greater in vertebrates due to an increased number of splice variants
and an increased number of genes, likely due to gene duplication events. AMPAR
subunits form an orthologous group, and there is high variability in the number of
AMPAR genes in each species due to extensive taxon specific gene gain and loss. qPCR
results show that all 12 Aplysia.iGluR subunits are expressed in all nervous system
ganglia. Orthologous NMDAR subunits in all species studied suggests conserved
function across Bilateria, and potentially a conserved mechanism of neuroplasticity and
learning. Our results suggest a significant role for L-Glu mediated responses throughout
15
16
the Aplysia nervous system, consistent with L-Glu’s role as the primary excitatory
neurotransmitter.
2.2 Background
L-Glu is the most abundant neurotransmitter in the vertebrate brain (Meldrum
2000), and exerts most of its effects by binding to different postsynaptic ligand-gated
receptors. L-Glu receptors are classified into two types, ionotropic and metabotropic
receptors (Petralia, et al. 1996). Metabotropic L-Glu receptors (mGluR) are G-proteincoupled receptors in which binding of L-Glu activates intracellular cascades and
modification of intracellular proteins. Ionotropic L-Glu receptors (iGluR) convey the
majority of fast excitatory signal transmission and have been implicated in most aspects
of central nervous system (CNS) development and function (Traynelis, et al. 2010b). The
binding of L-Glu to iGluR opens transmembrane ion channels that allow ions to cross the
plasma membrane leading to depolarization of the postsynaptic cell and triggering of
action potentials, thereby transmitting synaptic information. iGluR play important roles in
synaptic plasticity, which is the ability of a synapse to strengthen or weaken its
interactions with others over time in response to changes in activity. This feature of
iGluR is believed to be a key mechanism underlying learning and memory (Danysz, et al.
1995).
Several features of iGluR in vertebrates have been revealed using model systems
such as rats and mice. Vertebrate iGluR are divided into three subtypes according to
selective agonists: N-methyl-D-aspartate receptors (NMDAR), kainate, and α-Amino-3hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPAR). Each subtype of iGluR
17
is composed of 4 subunits that form a dimer of dimers to create the functional protein
(Tichelaar, et al. 2004). In a phylogenetic analysis of human, rat, and mouse iGluR
subunits, each of the 3 different subtypes formed a monophyletic clade, with functional
iGluR proteins made only with subunits within each individual clade (Lipsky and
Goldman 2003). Overall amino acid identity of iGluR subunits across the 3 subtypes in
mammals is only 20-30%, but all subtypes contain common structural features that place
them together into a single large superfamily (Kew and Kemp 2005). A fourth iGluR
subtype, called delta receptors, show low sequence similarity with other iGluR and do not
open ion channels, but are believed to bind D-serine and glycine (Naur, et al. 2007).
Invertebrate model species such as Aplysia californica (Aplysia), Drosophila
melanogaster (Drosophila), and Caenorhabditis elegans (C. elegans) have been used
extensively in studies of iGluR mediated transmission in the nervous system since the
1960’s due to several distinct advantages over vertebrate models (Kandel 2001). Their
nervous systems contain 302-135,000 neurons, compared to 1x1011 in the human brain
(Kosinski and Zaremba 2007; Herculano-Houzel 2009; Alivisatos, et al. 2012), and
simple neuronal circuits underlying various behaviors have been described (Walters, et
al. 1983a; Chalfie, et al. 1985; Simpson 2009; Piggott, et al. 2011). There is ample
evidence suggesting that L-Glu is the major neurotransmitter in many neural circuits in
these species (Dale and Kandel 1993; Levenson, Endo, et al. 2000; Brockie, et al. 2001;
Marrus, et al. 2004). These advantages have resulted in widespread use of Aplysia,
Drosophila, C. elegans, and other models to study iGluR mediated synaptic plasticity,
because learned behaviors can be correlated with both molecular and physiological
synaptic changes between sensory- and motoneurons.
18
Physiologically, invertebrate model organisms have greatly enhanced our
understanding of L-Glu mediated synaptic plasticity. Many plasticity related changes
characterized in invertebrate models subsequently have been demonstrated to occur in the
more complex vertebrate hippocampus (Malenka and Bear 2004). For example, AMPAR
and NMDAR have been implicated in learning and memory in vertebrates, Aplysia, and
D. malanogaster (Lisman, et al. 1998; Lee, et al. 2003; Xia, et al. 2005). Despite the
physiological use of invertebrate model species for iGluR-mediated responses, an
outstanding question remains as to which iGluR genes are likely to play the functional
role of vertebrate NMDAR and AMPAR during plasticity.
To address this question we have conducted a phylogenetic analysis spanning all
three Bilateria superclades in order to identify iGluR genes in invertebrate model species
that are orthologous with vertebrate iGluR, and thus more likely to be functionally
similar. Bilateria are organized into three superclades based on embryology, morphology
and molecular data: Ecdysozoa (including arthropods like D. melanogaster, nematodes
like C. elegans), Lophotrochozoa (including molluscs such as Aplysia), and
Deuterostomia (including chordates like rats, mice, humans) (Philippe, et al. 2005).
Lophotrochozoa and Ecdysozoa are sister-clades that together form the Protostomia
(Halanych 2004; Dunn, et al. 2008). Most studies support the monophyly of these three
superclades, and thus a common ancestor for all Bilateria. Several recent studies support
a monophyletic grouping of Deuterostomia, with Protostomia as a sister clade (Paps, et
al. 2009).
Identification of Protostomia iGluR genes homologous to vertebrate NMDAR and
AMPAR subunits can allow for predictions of subunits that may be involved in observed
19
synaptic plasticity. This analysis can also add useful information to our poor
understanding of subtype-specific agonists in protostomes (Trudeau and Castellucci
1993; Fox and Lloyd 1999; Kimura, et al. 2001).
Aplysia is a model organism with a long history of studies of iGluR-mediated
nervous system function, in particular for learning and memory paradigms (Bartsch, et al.
1995; Kandel 2001; Kempsell and Fieber 2015b, a). Aplysia NMDAR subunits have been
shown to be expressed throughout the nervous system (Ha, et al. 2006), however most
other Aplysia iGluR have been identified through similarity to sequences of other species,
and their in vivo expression patterns are unknown. Subunits within each iGluR subtype
form a monophyletic clade in vertebrates, and complete receptors can only be formed
with subunits within each of these clades. An Aplysia-only phylogeny was built to
identify subunits that form monophyletic clades, and thus may form functional receptors.
In this study we addressed the evolution of bilaterian iGluR, including Aplysia,
using phylogenetic analysis. The expression of iGluR throughout the Aplysia nervous
system and the number of iGluR genes in Aplysia were also addressed to place this model
into an appropriate context with other iGluR model species.
2.3 Materials and Methods
Phylogenetic analysis
iGluR sequences for phylogenetic analysis were obtained from NCBI for all
species except Branchiostoma belcheri (lancelet), which were obtained from the Chinese
Lancelet Genome Project (<genome.bucm.edu.cn/lancelet>). Sequences of all identified
canonical iGluR proteins were aligned in MEGA7 (Kumar, et al. 2016) using the
20
MUSCLE (multiple sequence comparison by log-expectation) multiple aligner algorithm
with default parameters (Edgar 2004). Poorly aligned regions of the alignment were
removed using trimAl (Capella-Gutiérrez, et al. 2009) with the –automated1 option to
heuristically determine optimal trimming of the alignment, resulting in 473 positions used
for phylogenetic analysis. Unrooted phylogenetic trees were then constructed in MEGA7
using maximum likelihood and 1000 bootstrap replicates with the LG amino acid
substitution model (Le and Gascuel 2008) with gamma distributed rates and 5 rate
categories. The initial tree was obtained by applying the Neighbor-Joining algorithm to a
matrix of pairwise distances estimated using a JTT model.
Subtype specific trees were built to further clarify the relationships within
orthologous groups using the same parameters as the full phylogeny, except 500
bootstrap replicates were performed. These same parameters were also used to build the
Aplysia-only iGluR tree. Trees were visualized using FigTree (v1.4.2) (Rambaut 2012).
Identification of iGluR genes in Aplysia
Previously described iGluR in both Aplysia and chordates were obtained from the
NCBI database, and analyzed using tools at the SMART (Letunic, et al. 2012) and
Interpro (Hunter, et al. 2011) protein databases for potential binding sites and
transmembrane domains to identify all iGluR subunits in Aplysia. These sites are likely to
be highly conserved to maintain L-Glu activation, thus their sequences were extensively
searched in both the Aplysia published genome (NCBI) and the freely accessible Aplysia
transcriptome database (<http://www.aplysiagenetools.org/>). Candidate genes were then
translated to protein sequences and run through a BLAST search as well as scanned in
Interpro for verification. Genes were confirmed to be in the Aplysia transcriptome by
21
PCR amplification from cDNA of the abdominal ganglia and cloned in a TA vector
(Invitrogen) and subsequently sequenced.
Hydrophobicity analysis
The Kite/Doolittle hydrophobicity scale was used to determine the hydrophobicity
of each amino acid for the last 450 amino acids in both human and Aplysia Grin1
subunits, which contains the ligand binding, transmembrane, and intracellular C-terminal
domains. For representative AMPAR and kainate receptor subunits only transmembrane
domains were analyzed due to lack of agonist-binding site identification in Aplysia.
mRNA extraction and quantification of iGluR gene expression in the nervous system
To describe quantitative expression of identified iGluR in Aplysia nervous system
ganglia, six sexually mature Aplysia californica from a single egg mass of wild-caught
animals were obtained from the National Resource for Aplysia at the University of
Miami. Animals were anesthetized in a solution of 50:50 isotonic MgCl2:Artificial sea
water (ASW). All ganglia were removed, immediately rinsed in ASW, and placed in
Trizol (Invitrogen). Tissues were ground in a bead homogenizer to break cells out of the
sheath prior to RNA extraction. Both hemiganglia for each tissue from each animal were
pooled into a single sample.
Total RNA was extracted following the manufacturer’s protocol and samples
were treated with DNAse to remove any contaminating DNA. RNA quantities were
determined using a Nanodrop (Model ND-1000), and samples were stored at -80°C until
further processing. 100 ng of RNA was reverse transcribed into cDNA using the
SuperScript III First-Strand Synthesis System (Invitrogen). After dilution of the cDNA
(1:5 with H2O) messenger RNA (mRNA) copy number was determined using qPCR on a
22
Stratagene Mx3005P with SYBR Green master mix and the equivalent of 2 ng of starting
RNA per well.
Primer pairs were designed for each of the thirteen iGluR identified in Aplysia to
detect expression levels with quantitative real-time PCR (Table 2.1). Efficiencies of each
primer pair was tested by generating standard curves based on regression analyses of the
Ct and the log value of 10-time dilution of each target gene for each primer pair. All
primers used had efficiencies between 0.9-1.1. mRNA copies were then calculated using
standard curves and the average of duplicate cycle threshold (Ct) values.
Southern blot
Southern blotting from agarose gels was done onto Hybond-N+membrane
(Amersham) in sodium saline citrate buffer (10×SSC; 1.5 M NaCl; 0.15 M Na citrate pH
7.0) by capillary action. Hybridization for high stringency blots was conducted in 30%
formamide, 5× SSC, 1× Denhardt’s, 0.2 % sodium dodecyl sulphate (SDS), 10% Dextran
sulphate, 20 mM sodium phosphate, pH 6.8 and at 42°C. Final washes for high
stringency were in 0.1 ×SSC, 0.1 % SDS at 65°C. Each lane contained 4µg of DNA and
was cut with one of the following enzymes: EcoR1, HindIII, BamH1, and PstI. The probe
was labeled using a random primed DNA labeling kit with [g-32P] dCTP.
2.4 Results
Phylogenetic analysis of bilaterian iGluR subunits
To investigate the evolutionary relationships of iGluR across the Bilateria,
phylogenetic analysis was conducted on full-length protein sequences of all NMDAR,
AMPAR, kainate receptor, and delta receptor subunits. Sequences were included from the
23
three major bilaterian lineages: Deuterostomia (Homo sapiens, Mus musculus, Rattus
norvegicus, Danio rerio, Branchiostoma lanceolatum, Ciona intestinalis), Ecdysozoa
(Limulus polyphemus, Priapulus caudatus, Drosophila melanogaster, Caenorhabditis
elegans, Daphnia magna, Tribolium castaneum), and Lophotrochozoa (Aplysia
californica, Octopus bimaculoides, Lingula anatina). The species tree showing the
evolutionary relationships among the species is shown in Fig. 2.1.
The phylogenetic tree of bilaterian iGluR is presented in Fig. 2.2. Identification of
iGluR genes as kainate receptor or AMPAR subtypes is unclear for many protostomes,
thus these genes are currently named as “GluR” genes without an AMPA or kainate
designation. Deuterostome iGluR genes are named Grin (Glutamate Receptor Ionotropic
NMDA), GRIA (Glutamate Receptor Ionotropic AMPA), or GRIK (Glutamate Receptor
Ionotropic Kainate) based on selective agonists. Canonical sequences, as determined by
Uniprot, were used for genes with more than 1 splice variant. Some deep branches of the
tree were poorly supported by bootstrapping due to high divergence of sequences
between subtypes and large evolutionary distance between species. However, most
orthologous groups were well resolved with strong bootstrap support. Based on the
phylogeny of the iGluR proteins each protostome subunit was classified as NMDAR,
AMPAR, kainate receptor or as orphan genes that do not show a clear relationship with
vertebrate subtypes (Table 2.2).
NMDAR genes form three orthologous groups corresponding to Grin1, Grin2,
and Grin3 subunits, providing evidence that all three ancestral Grin subunits were present
in the most recent bilaterian ancestor (Fig. 2.2). This is a unique feature of NMDAR: they
are the only iGluR subtype with more than one orthologous copy present before the
24
divergence of protostomes and deuterostomes. In the orthologous group Grin2 each
protostome and basal deuterostome has only one ortholog of Grin2, but vertebrates have
four Grin2 genes. Vertebrate Grin2 genes form a highly discrete clade, with four
paralogous copies that arose early in the vertebrate lineage. The Grin2 paralogs in
vertebrates are likely best explained by the 2R hypothesis, which postulates that two
rounds of whole genome duplication occurred early in the vertebrate lineage after their
split from tunicates (Dehal and Boore 2005; Putnam, et al. 2008). Thus, the 2R
hypothesis predicts that vertebrates are expected to have four copies of each gene in
comparison to one copy in invertebrates. In the case of Grin2 all four paralogs have been
retained in all vertebrates used in this study.
In contrast, both in Grin1 and Grin3 there is only a single orthologous group of
vertebrates present in the tree. Therefore, in both cases only one of the four that
originated during the 2R genome duplications remained active, whereas three of them
have been lost early in vertebrate evolution. Despite large evolutionary distances among
the studied species, all three Grin orthologs are highly conserved suggesting that they are
under high functional constraints slowing their divergence.
For AMPAR genes there is a monophyletic clade of all vertebrate AMPAR
sequences and several orthologous protostome genes, pointing to an AMPAR gene copy
present in the common bilaterian ancestor (Fig. 2.2). The number of AMPAR genes in
protostome species is highly variable and appears to be species or taxon specific. For
example, in the Lophotrochozoa, Aplysia has 6 paralogous AMPAR genes, Octopus has
1, and Lingula has 2. Ecdysozoan species have 2-3 genes in this orthogroup, except
Daphnia and Tribolium, which do not have any genes in the AMPAR orthogroup.
25
Most AMPAR genes of protostome species are most closely related to each other
(inparalogs) rather than to AMPAR genes, with Lingula as the exception. This implies
that gene duplications occurred after the divergence of each lineage in this study,
suggesting that there has been extensive gene gain and loss that has acted independently
in each protostome taxon. This appears particularly true in Aplysia. Six AMPAR genes in
Aplysia and 1-2 in other lophotrocozoans, including Octopus, suggests several gene
duplication events occurred within the Gastropoda lineage. The sampled vertebrates have
consistent with the 2R genome duplication scenario all four AMPAR genes, in
comparison to one AMPAR ortholog in the common bilaterian ancestor.
Kainate receptor genes are the only subtype in the tree in which Protostomia and
Deuterostomia genes do not form a strongly supported monophyletic clade. Only three
protostome genes form an orthologous group with chordate kainate receptor subunits:
Aplysia KA2, Tribolium GRIK1, and Limulus GRIK2. Identification of kainate receptor
genes is unclear for many Ecdysozoan species due to extensive divergence of these
subunits from Deuterostomia. Many predicted kainate receptor genes from Limulus,
Priapulus, and Tribolium form an independent clade without a clear relationship to
chordate kainate receptor genes (to the left of the tree, Fig. 2.2). This suggests they have a
slightly different function, or that they work through a different mechanism than the
kainate genes that have been studied. Additionally, Priapulus, Limulus, Daphnia, and
Tribolium have several genes that appear to be very distant to all other genes in the tree
(to the right of the tree, Fig. 2.2).
Within vertebrates there are two ancient paralogs with a total of five kainate
genes. The three paralogs GRIK_A1-3 form one orthologous type and the two paralogs
26
GRIK_B4-5 form together with the lancelet GRIK2 sequence the second type. Therefore,
in comparison to the four paralogs expected under the 2R duplications, GRIK_A1-3 and
GRIK_B4-5 were respectively losing one and two paralogs. In contrast to Grin and
AMPAR genes, extensive divergence of kainate receptor genes in Deuterostomia and
Protostomia suggests that they are the least conserved iGluR subtype across Bilateria.
Orthologous groups of each iGluR subtype were further analyzed in the attempt to
obtain a better resolution of the evolutionary relationships in these parts of the tree
(Supplementary Figs. S2.1-S2.3, pg. 136-138). In this analysis the relationship between
protostome and chordate iGluR were the same as in the full phylogeny, providing further
support for the relationships found in the full phylogeny.
An additional phylogenetic tree was built using only Aplysia iGluR to search for
subunits that form monophyletic clades and thus may form complete receptors
(Supplementary Fig. S2.4, pg. 139). The Aplysia-only tree confirms the findings of the
full phylogeny, with subunits corresponding to currently predicted NMDAR and
AMPAR subunits forming monophyletic clades, while the four predicted kainate receptor
subunits do not show a monophyletic relationship.
Number of iGluR genes in Aplysia and other bilaterians
Through genomic searches, the number of iGluR genes identified in chordates
was greater than the number of genes in any protostome (Table 2.3), although it must be
noted that for some protostome species the number of iGluR genes may not be accurate
due to limited availability of genomic information and annotation. As described in the
phylogenetic analysis, extensive gene gain and loss of AMPAR genes in protostomes has
27
also contributed to the variable number of iGluR genes in different protostome species.
The increased number of genes in chordates is likely due to retention of many paralogs
after 2R genome duplications. We found that in both Grin1 and Grin3 orthologous groups
three of the four paralogous genes from the 2R genome duplications have not been
retained.
Genomic searches using SMART (Letunic, et al. 2015) and InterPro tools
(Mitchell, et al. 2014) revealed 12 unique iGluR genes and one splice variant of Grin1 in
Aplysia. The translated Aplysia iGluR sequences showed a highly conserved
SYTANLAAF motif (Wollmuth and Sobolevsky 2004) near the second transmembrane
loop, which was used to identify them as iGluR (Fig. 2.3). This motif contributes to
formation of the channel outer pore and its activation gate (Dai and Zhou 2013), and
amino acid substitutions in this motif are reported to alter channel gating and
permeability (Kohda, et al. 2000; Low, et al. 2003). Aplysia showed a greater number of
iGluR subunits compared to other protostomes, primarily due to several gene duplications
of AMPAR genes, as discussed above. cDNAs corresponding to all 12 Aplysia iGluR
were isolated from the nervous system and confirmed to be transcribed in vivo.
Sequence similarity and conserved domains of Grin1 in Aplysia and vertebrates
Conserved domains of Grin1 proteins in Aplysia and the vertebrates studied were
analyzed to elucidate whether similarities in these parts of the sequences might predict
conservation of function. The first ~400 amino acids of NMDAR proteins, known as the
N-terminal domain (NTD), show high sequence divergence, with only 19% sequence
identity in this region within a species. Mutant Grin2 subunits lacking the entire NTD
were shown to assemble into receptors functionally similar to complete receptors, thus it
28
is not surprising that this region showed low sequence similarity between species
(Fayyazuddin, et al. 2000; Gielen, et al. 2009). Comparatively, agonist binding domains
(ABDs) and transmembrane domains (TMDs) must show greater conservation to
maintain iGluR functionality, and indeed, within a species these regions show 63% and
73% identity, respectively (Paoletti 2011). In comparing the Grin1 subunits of H. sapiens
and Aplysia we found 66% sequence identity between these subunits after removal of the
NTD, similar to the sequence similarity of ABDs and TMDs within a species.
The three TMDs of Grin1 together form the iGluR ion channel and these regions
show very high protein sequence similarity throughout the Bilateria. However, Aplysia
sequences consistently showed greater sequence similarity to H. sapiens than did either
D. melanogaster or C. elegans (Table 2.4). The TMDs of the vertebrate genes compared
had close to 100% conservation, on average, and exhibited much greater amino acid
similarity than the rest of the protein, suggesting high selective pressure on these sites to
maintain function.
The hydrophobicity of amino acids in a protein influences the folding and
structure of the molecule, and protein regions with similar hydrophobicity profiles are
predicted to maintain structural stability. TMDs are well conserved, but substitutions that
result in different amino acids occur, as shown in Table 2.4. A plot of the hydrophobicity
of the canonical Grin1 sequences in Aplysia and H. sapiens shows that amino acid
substitutions in the TMD’s have been tolerated only when the substituted residue has a
similar hydrophobicity, and thus is predicted not to significantly alter protein folding
(Fig. 2.4). This result suggests that the structure and function of the transmembrane
domains and ion channel have been maintained. Conversely, protein sequences for the
29
glycine and NMDA binding sites show high sequence divergence and numerous changes
in hydrophobicity compared to H. sapiens sequences. Hydrophobicity plays a crucial role
in receptor binding domains (De Loof, et al. 1986), and this lack of conservation may
reflect the diminished role for glycine as a co-agonist in Aplysia NMDAR (Carlson, et al.
2012).
Representative hydrophobicity plots of AMPAR and kainate receptor subunits can
be found in Supplementary Figs. S2.5-S2.6, pgs. 140-141. For both AMPAR and kainate
receptor subunits, hydrophobicity of the TMDs appears to be well conserved with H.
sapiens, however, they are not as conserved as the TMDs of Grin1 subunits. This
supports the phylogenetic evidence that AMPAR subunits in Aplysia are closely related
to H. sapiens subunits and may perform similar functions. In the phylogeny, kainate
receptor subunits of the prostostomes were highly divergent from vertebrates, however
the TMDs are a notable exception and appeared to be well-conserved with H. sapiens in
this analysis.
Evaluation of Grin2 genes in Aplysia genomic DNA
While a search of the recently revised A. californica genome revealed only one
gene similar to Grin2 subunits in Chordata, the presence of gaps in this genome sequence
raises the possibility that additional genes may have been missed. A Southern blot
analysis of genomic DNA of Aplysia demonstrated only a single band homologous to a
conserved region of Grin2 (data not shown). This band was at the expected molecular
weight for the previously identified Aplysia Grin2, suggesting that no other Grin2 genes
are present in this genome. This result was not surprising considering that there is only 1
Grin2 gene found in other Protostomia species (Teng, et al. 2010).
30
Quantification of iGluR gene expression in the nervous system
The expression of identified iGluR in all Aplysia nervous system ganglia is shown
in Fig. 2.5. All twelve subunits, as well as the splice variant Grin1-2, were expressed in
all ganglia, with differences both within and between different iGluR subtypes. The
subtype with the highest expression varied across the different ganglia (Fig. 2.5A). In the
pleural and buccal ganglion, kainate-like receptor genes were expressed at the highest
levels, followed by NMDAR-like genes, with AMPAR-like genes expressed at the lowest
levels. The highest expression of all 3 iGluR subtypes was found in the pedal ganglion,
suggesting that this ganglion has the greatest reliance on L-Glu-mediated
neurotransmission. The buccal ganglion showed the lowest expression for both NMDAR
and AMPAR genes and showed significantly greater (p£0.05) expression of kainate
receptor subunits.
Expression of NMDAR genes was primarily due to high expression of Grin1-1,
representing >60% of NMDAR expression in nearly all nervous system tissues (Fig.
2.5B). These data suggest that Grin1 may be the subunit that contributes the most to
NMDAR-mediated physiological responses to L-Glu in the nervous system of Aplysia.
The lone exception was the buccal ganglion, where the splice variant Grin1-2 was the
most highly expressed NMDAR subunit gene in the majority of the 6 animals studied. A
similar pattern of high expression of 1 gene was observed with kainate receptor genes,
where the GluR7 subunit comprised the majority of this subtype’s gene expression. The
distribution of expression of AMPAR subunits favored no single subunit gene
dominating the expression of this subtype.
31
2.5 Discussion
iGluR mediated responses in the nervous system are of particular interest due to
their role in synaptic plasticity associated with learning and memory. Protostomes have a
long history of use as model organisms for studies of iGluR mediated synaptic plasticity
due to their simple nervous systems and well-defined neural circuits. Many plasticityrelated changes in the vertebrate nervous system were first discovered in protostomes,
and were subsequently shown to be conserved in chordates (Kandel 2001). Further
studies in vertebrates have identified the NMDAR and AMPAR subtypes of iGluR
underlying these processes. Despite the utility of protostomes for iGluR mediated
responses in the nervous system, identification of iGluR subunits that are orthologous to
chordate iGluR has so far not been thoroughly studied. The application of phylogenetic
methods to specific protein families is a useful procedure to shed light on their functional
status.
We conducted a phylogenetic analysis of iGluR spanning all three superclades of
bilaterians; Deuterostomia (including Chordata), and their sisterclade Protostomia,
consisting of Ecdysozoa and Lophotrochozoa. We found that the evolutionary history of
iGluR in Bilateria is distinct for each iGluR subtype. It was possible to identify
protostome iGluR subunits most closely related to chordate iGluR, with possibly similar
functional roles, for both NMDAR and AMPAR genes.
The best supported conclusion from the analysis presented is that clear
orthologous groups are formed for both Grin1 and Grin2 genes. This is an extension of
the previous report that protostome Grin1 and Grin2, including Aplysia, are more closely
related to chordate Grin1 and Grin2, respectively, than to each other (Ha, et al. 2006) and
32
further supports the close kinship of all known bilaterian NMDAR subunits. This study
confirms an ancient duplication that generated two orthologous NMDAR genes present in
the common bilaterian ancestor. Both orthogroups have been under high selective
constraint that has maintained the function of these genes at least since the divergence of
deuterostomes and protostomes, thus making protostome model organisms excellent
study systems for NMDAR.
The number of NMDAR subunits varies greatly between deuterostomes and
protostomes. Chordate species in this study express 3-8 Grin1 splice variants, while only
1 Grin1 splice variant was identified in Aplysia. The chordates also express a greater
number of Grin2 genes and splice variants. Two independent rounds of whole genome
duplication in the deuterostome lineage since its split from the bilaterian ancestor, that
also gave rise to Lophotrochozoa (Dehal and Boore 2005), are hypothesized to be the
origin of these duplicated Grin2 genes (Teng, et al. 2010; Ohno 2013). The regulatory Cterminus contains most of the sequence variability in different vertebrate Grin2 genes.
We hypothesize that the retention of duplicated Grin2 genes is explained by the
subfunctionalization model, whereby each duplicate gene copy maintains a subset of its
original function (Force, et al. 1999).
The NMDAR subunits involved in receptor assembly have been shown to
strongly alter pharmacological and biophysical properties of the channel, including
sensitivity to allosteric modulators, single channel conductance, and
activation/deactivation kinetics (Stern, et al. 1992; Rumbaugh, et al. 2000; Cull-Candy
and Leszkiewicz 2004). The greater number of NMDAR subunits available for complete
receptors in chordates increases the different physiological responses possible, and likely
33
underlies the more complex and nuanced signaling capability observed in vertebrates.
Furthermore, activation of NMDAR has been shown to be an essential component of
synaptic plasticity and memory formation in vertebrates (Hunt and Castillo 2012), and
knockdown of NMDAR can disrupt learning in protostome species (Xia, et al. 2005;
Esdin, et al. 2010; Müßig, et al. 2010). In spite of the greater diversity of NMDAR
subunits in chordates, homology of Grin1 and Grin2 across bilaterian animals suggests
that a wealth of discovery in the mechanisms of NMDAR-induced plasticity is
nevertheless possible in simpler nervous systems such as Aplysia, with relevance to
vertebrate NMDAR physiology, including that of humans.
AMPAR subtype genes also form an orthologous group, with evidence of a single
ancestral AMPAR gene in the common bilaterian ancestor. Unlike NMDAR, the number
of protostome genes in this orthogroup is highly variable and has been subject to
extensive lineage and taxon specific gene duplications. For example, in the Ecdysozoan
superclade, Pripaulus, Drosophila, and C. elegans each have two inparalogous AMPAR
genes most closely related to genes in their own species, suggesting that each gene is a
paralog that arose independently in each lineage. Interestingly, the other two Ecdysozoan
species studied, Daphnia and Tribolium, do not have any genes within the AMPAR
orthogroup. In Lophotrochozoans expansion of AMPAR genes is particularly prevalent in
Aplysia, where multiple gene duplication events have resulted in six AMPAR genes.
Whether AMPAR physiology in protostomes will bear the same resemblance to
chordate AMPAR, as is the case with NMDAR, is uncertain. The name AMPAR for this
subtype is likely inaccurate in protostomes. Isolated Aplysia sensory neurons do not
respond to exogenously applied AMPA, and AMPAR antagonists do not block whole cell
34
currents elicited by the L-Glu analog D-aspartate (D-Asp; (Carlson, et al. 2012)). Yet
Aplysia AMPAR have physiological roles with relevance to vertebrate learning, with
AMPAR antagonists inhibiting facilitation (Chitwood, et al. 2001; Li, et al. 2005).
Synaptic plasticity in vertebrates is also AMPAR-dependent (Zamanillo, et al. 1999;
Sanderson, et al. 2008). AMPAR antagonists such as 6,7-dinitroquinoxaline-2,3-dione
disodium (DNQX) and 6-cyano-7-nitroquinoxaline-2,3-dione disodium (CNQX) act on
sites that are different than the agonist binding site. Similarity to chordate AMPAR
suggests that Aplysia and other protostome AMPAR may bind agonists and antagonists
differently than do those of vertebrates, yet operate similarly physiologically once
activated.
Extensive gene gain and loss of AMPAR suggests less functional constraint
compared to NMDAR genes, and may underlie these dissimilarities in AMPAR
physiology in different taxonomic groups. Different numbers of AMPAR genes in
different protostomes may result in unique receptor assemblages conveying different
physiological properties in each taxon, thus the exact role of AMPAR may be unique in
different lineages.
Kainate receptors are the least conserved subtype, with no strongly supported
monophyletic group, but were clearly present in the common bilaterian ancestor. Kainate
receptors play a minor role in synaptic signaling (Song and Huganir 2002), and are not
believed to be as involved in learning and memory as NMDAR and AMPAR (Contractor,
et al. 2000). Therefore, it is unsurprising that there may be less functional constraint on
these receptors than there is on AMPAR and NMDAR. Large divergence of kainate
35
receptor genes makes it difficult to predict the functional relevance of these genes to
iGluR mediated excitability in the nervous system of protostomes.
In agreement with previous studies, vertebrate iGluR subunits were arranged into
three distinct clades corresponding to the 3 different agonists (Lipsky and Goldman
2003). Our analysis of Aplysia only iGluR genes indicated that NMDAR and AMPAR
subunits form well-defined clades and thus may be capable of forming functional iGluR
channels. In contrast, the makeup of kainate receptors is less clear, due to the lack of
monophyletic groups in the protostomes.
All 12 identified iGluR subunits were expressed in all Aplysia ganglia, extending
the results of an in situ hybridization study concluding that Grin1 was expressed
throughout the nervous system (Ha, et al. 2006). This attests to the importance of L-Glu
mediated fast synaptic transmission in all parts of the Aplysia nervous system. Grin1
expression was highest in nearly all Aplysia ganglia, with its two splice variants Grin1-1
and Grin1-2 together comprising ~75% of total NMDAR expression. A study using a
nuclease protection assay showed that expression of Grin1 comprised 67-88% of the total
Grin expression in rat brain (Goebel and Poosch 1999). When combined with homology
discussed earlier, this suggests that the regulation and function of NMDAR in Aplysia are
highly conserved with those of vertebrates.
Variable and spatially distinct expression of the three iGluR subtypes was
observed in the ganglia of the Aplysia brain. Pedal ganglion had the highest iGluR
expression, with the glutamatergic nature of pedal motoneuron transmission corroborated
by physiological studies. (Walters, et al. 1983a, b). Variations in the frequency and
amplitude of ionic currents activated by the iGluR agonists L-Glu and D-Asp have been
36
documented in neurons isolated from different ganglia (Carlson and Fieber 2011;
Kempsell and Fieber 2014), lending support to the non-uniformity of the receptor
expression patterns. Studies in mammalian brains have shown both spatial and
developmental variations in patterns of expression of NMDAR and AMPAR subunits,
with some subunits specific for certain brain regions, or variable expression dependent on
the stage of development (Akazawa, et al. 1994; Mansour, et al. 2001).
To place bilaterian iGluR into a larger evolutionary perspective, insights about the
deep origins of iGluR have recently emerged from discoveries on ctenophores. Studies of
candidate iGluR genes in the ctenophore Mnemiopsis leidyi, which is currently thought
basal to bilaterian animals (Moroz, et al. 2014), revealed that ctenophore iGluR form a
monophyletic clade separate from, and ancestral to, chordates (Ryan, et al. 2013). Thus it
appears that subunit types emerged after the ctenophores split, but before the divergence
of Deuterostomia, Ecdysozoa, and Lophotrochozoa. This is suggestive of individual
AMPAR and NMDAR subunits evolving prior to the last common bilaterian ancestor,
but after divergence from ctenophores. Furthermore, two studies using expressed
sequence tags (EST), genome organization, gene structure and functional content found
lower amino acid substitution rates in Lophotrochozoa than Ecdysozoa relative to
chordates. These findings suggest that genes in Lophotrochozoa are more likely to have
greater sequence similarity to chordates than Ecdysozoan genes, and hence may be more
likely to be functionally equivalent to chordates.
NMDAR and AMPAR subtypes of iGluR are vital to synaptic plasticity
associated with vertebrate learning. This study confirms the ancestral origins of NMDAR
and AMPAR genes and also, but less strongly supported, ancestral kainate receptor genes
37
in Bilateria. These findings underscore the utility of Aplysia and other protostome models
for the studies of AMPAR and NMDAR mediated responses in the nervous system.
This is the first analysis of the phylogenetic relationships between subtypes of
iGluR genes across Bilateria. For decades, model organisms from the Protostomia have
been used as models of nervous system function, and we show that AMPAR and
NMDAR subtypes were present in the common bilaterian ancestor and have been
maintained as orthologous groups. Functional constraint preventing amino acid
substitutions in pore regions of NMDAR suggests a highly conserved function of these
subunits and potentially a conserved mechanism of learning. Kainate receptor subunits
are the least conserved and may not play the same role in protostomes and deuterostomes.
qPCR results demonstrate that iGluR are expressed ubiquitously throughout the nervous
system of Aplysia, underscoring the importance of this model to understanding iGluR
mediated nervous system function.
38
Table 2.1 Primers used for qPCR.
Gene
Primer Sequences for qPCR (5’-3’)
GluR1
F
R
F
R
F
R
F
R
F
R
F
R
F
R
F
R
F
R
F
R
F
R
F
R
F
R
GluR2
GluR3
GluR4
GluR5
GluR6
GluR7
GluR8
KA1
KA2
Grin1-1
Grin1-2
Grin2
Amplicon length (bp)
GCTTTGTGGACAACACCAGC
GGTTCTGCCATGATGATCGAC
116
GACTTAAAGGTGTCCAACGC
CGTCATACGACAAGCTCTTC
GAACCTTGACCCCAAGTTCTG
CCTATTCACGAGAGCTTTCG
CGTACTTGAGTTTGCGGTCC
GTGACCATATCGCTGCAGAC
CTTCACGGAGGAGTCAAAGTC
CTTCAGACGGATGCAGCACT
GAGATACGTGCTGGATCAGG
GGCTCATGATGGACTACAAC
CGAGACTGCTTTAGCCTACG
CCATTAGCCTGCCTTGTGAC
TCAAGTCGCTCAATCTCTCC
CCACGTCAACACACTTCTAGTC
GCTACCGAAACCAGCTCAATC
GCAGTCACACGTACTTAAGAGG
GATCTGGCTCTACGTCATTGG
CTTGTAACCCAGACACGGAC
GGGAAACATTGAGAAGC
GCTCCCAATGCAAACACAGC
145
GACACCAACGAAATAGACCTC
GCTCCCAATGCAAACACAGC
AGTTCACCTGCGACTCTGAC
CCGAGGTTCTCATCCTGAATC
146
101
121
146
154
100
110
145
115
135
150
39
Table. 2.2. Placement of Protostomia iGluR into subtypes based on phylogenies.
Species
Aplysia californica
AMPA
GluR1
GluR2
GluR3
GluR4
GluR5
GluR8
GluR1
GluR2
Kainate
GluR7
KA2
NMDA
Grin1
Grin2
Orphans
KA1
GluR6
none
Grin1
Grin2
Drosophila
melanogaster
GluR1A
GluR1B
GluR1C
GluR2D
GluRIIE
Grin1
Grin2
GluR3
GluR4a
GluR5
GluR6
GluR7
GluRIIA
GluRIIB
GluRIIC
Daphnia magna
none
GRIK1
Limulus polyphemus
GluR1
GRIK2
GRIK3
Lingula anatina
GluR2
GluR4
GluR
GRIK2
Priapulus caudatus
GluR3
GluR4
none
Grin1
GluR1
Grin2
Grin3
Grin1
Grin2
Grin3
Grin1
Grin2
Grin1
Grin3
Grin1
Grin2
Grin3
Tribolium castaneum
none
GRIK1
GluR1
Caenorhabditis
elegans
Octopus bimaculoides
none
Grin1
Grin2
GRIK2
GRIK1
GluR4
GluR2
GluR1
GRIK2
GRIK3
GluR1
GluR2
GRIK4
GRIK2
GRIK3
GRIK5
GluR2
iGluR genes in each protostome species were categorized into subtypes based on their
phylogenetic relationship with chordate iGluR genes. Many protostome iGluR genes do
not have a clear relationship with chordate subtypes and are thus identified as orphan
receptors. Protostome orphan receptors are divergent from chordate genes and thus
unlikely to perform the same subtype specific functions.
40
Table 2.3. Number of iGluR genes in the protostomes and vertebrates.
Species
Total iGluR
genes
Homo sapiens
14
Caenorhabditis
elegans
9
Mus musculus
14
Ciona intestinalis
6
Rattus norvegicus
14
Daphnia magna
7
Danio rerio
14
Limulus Polyphemus
11
Aplysia california
12
Priapulus caudatus
11
Lingula anatina
6
Tribolium castaneum
10
6
Branchiostoma
belcheri
10
Octopus
bimaculoides
Drosophila
melanogaster
10
Species
Total iGluR
genes
41
Table 2.4. Sequence similarities in transmembrane domains of Grin1 between
different protostomes and H. sapiens.
Species
A. californica
TMD1 TMD2 TMD3
85%
100% 87%
D. melanogaster 75%
100%
61%
C. elegans
87%
48%
65%
Compared to H. sapiens, Aplysia had fewer amino acid substitutions in transmembrane
domains (TMDs) than did D. melanogaster and C. elegans. This points to a higher
likelihood of conserved channel function between vertebrates and Aplysia.
42
Rattus
norvegicus
Rattus
norvegicus
Mus
musculus
Mus
musculus
Chordata
Homo
sapiens
Homo
sapiens
Deuterostomia
Danio
Danio
reriorerio
Branchiostoma
lancelolatum
Branchiostoma lanceolatum
Ciona
intestinalis
Ciona intestinalis
Lingula
anatina
Lingula
anatina
Lophotrocozoa
Cephalopoda
Mollusca
Gastropoda
Octopus
bimaculoides
Octopus
bimaculoides
Aplysia
californica
Aplysia californica
Priapulus
caudatus
Priapulus caudatus
Daphnia
Daphnia
magnamagna
Protostomia
Drosophila
melanogaster
Drosophila melanogaster
Arthropoda
Ecdysozoa
Tribolium
castaneum
Tribolium
castaneum
Limulus
polyphemus
Limulus
polyphemus
Nematoda
Caenorhabditis
Caenorhabditis
elegans elegans
3.0
Fig. 2.1. Phylogeny of species used for iGluR analysis. Tree of the evolutionary
relationships between species used in this study. The tree was built using the NCBI
taxonomy browser (Sayers, et al. 2011).
43
NMDA
Kainate
Grin2 _Dro soph ila
1
m
Gr in2 _T rib oliu
ib ol iu m
Gr in 1_ Tr
im ul us
G ri n3 _L
ia
n
_Daph
G ri n 3
pus
_ O c to
t
G ri n 3
cele
n
h
lu s
_La
is
A
f
ri a p u
rin3
e
bra
G
e
3_P
Z
ous
G ri n
3A_
_M
t
ia
G r in r in 3 A A_Ra
hn a
G
3
ns ap n i
n
ga _D p h l a s
ma
Grin
H u Cele luR1 _ D a p h i u l u lus
A_
_
G rin1 o s o L i m pu pus
i n 3 rin1
Gr
G _ D r 1 _ ria to a
G
in P c i
i n 1 G r 1_ _O lys
rin 1 p
G rin _A
G n1
ri
G
Grin2_Daphnia
1
0 .9 9
an
um se
_H ou
D
M
2 _
rin D
G rin2
lus
s
pu
G l a let
ulu
ria
u e
iap 1 _ P
r
n g anc
i
a
P
_ L _L
2_ D e l t u s
uR
R1 4 t
ul
l u IK le G l
us
iap
G GR nce
P r iapul
_
a
2
L
a
Pr
t
_
l
_
e
4
s
D RIK
K3
lu
I
u
G
GR
_Lim s
R2
lu
Glu _ L im u
1
a
lt
D e pulus
Pria nia
h
n ia
R1_
Glu 2_Dap 2 _ D a p h
GRIK D e lt a L im u lu s
2_
a
lt
e
D
m
pus
Tr ib ol iu
_ O ct o
D el ta 1_
D e lt a 1
ibo liu m
GR IK 5_ Tr
Glu R2 _Tr ibo lium
Delta2 _Tribo lium
GluRIC_Drosophila
GluRIB_Drosophila
Glu R1A _Dr oso
phi la
GluR1_Lim
ulus
GluR3_
P
GluR4 riapulus
_
GluR
_O Priapulus
GluR ctopus
3_Ap
ly
GluR sia
4_A
Glu
p
R2_ GluR1_ lysia
A
A
Glu Gl GluR plysia plysia
uR 5_A
R
GR
Gl 4_L 8_A plys
uR ing ply
ia
G GR IA
sia
u
2
G RIAIA1 1_M
Gl _ L i n l a
R 1 _R o
uR g u
I A _ a us
1_ l a
1 _ Hu t e
Gl Ce
Ze m
uR le
b r an
2_ ga
af
Ce ns
is
leg
h
an
s
se
ou
M at an ish
2_ R u m r a f e s h
IA _ H b s f i
R A2 _ e ou ra
G RI IA2 2_Z _M Zeban
G R I A A3 3 _ m
G R RI I A H u
ish
G G R _ at
raf h
3
G
IA _R Zeb rafis
GRRIA3IA4_ _Zeb
G G R A4B t
a
I
n
GRIA4_RHuma e
GRRIA4__Mous
G RIA4
a
G 1_Cionphnia
ulus
GluR 1_Da
3_Lim
GRIK
GRIK
liu m R 2D _D ro so ph ila
_ T ri b o
G lu
G lu R 1
rosophila
GluRIIE_D
GluR7_Aplysia
GRIK2_Lingula
G GR
RI IK
K
GR 1 _ 2 _ Z
IK Z e b e b
GR G
1
r
IK RIK GR _M rafi afis
3_ 1 IK o sh h
_
GR Zeb Hu 1_Ruse
IK3 raf ma at
GR
IK2 GR GR _M ish n
_La IK IK ou
nc 3_ 3_ s
GR GRIK elet Hum Rate
an
GRIKIK5_M5_Rat
GRIK
5 _ H ouse
GRIK 5_Zebrauman
4_Ze
fi
s
brafis h
h
GR
GRIK4_IK4_Rat
Mouse
GRIK4_Hu
man
KA2_Aplysia
GRIK 1_Tr iboliu m
GRIK2_Limulus
0 .8
0 .9 9
Gr
G
r
G in1
rin _L
1 in
G _Ze gu
rin b la
1 ra
Gr G _Hu fish
ri
i
Gl
u R GR n 1 _ n 1 _ m a
2 _ IA M o R a n
GR C i o 2 _ u s t
G IK2 n a L a n c e
G R R I K 2 _Cio
ele
I
_ T na
K
t
GR
1
IK3 _Lim ribo
_P
ulu lium
G R IK GluR7 riapulu s
_C
s
3_
GluR
4_Lim T r ib o li elegan
s
ulus u m
D e lt a
2_Oc
to p u
Delt s
D el ta 1_a1_Rat
M ou se
De lta 2_ Hu
De lta 1_ Ze bra m an
fis h
Delt a2_Z ebra fish
Delta1_Hum an
1
Delta2_Rat 1
Delt a2_M ouse
na
De lta 2_ Cio a
on
et
De lta 1_ Cielta1_Lancel
D
celet
1_Lan
GRIK ancelet
1_L
t
s
GRIA Lancele ctopu sia
a2_ RIK2_O1_Aply
Delt
G
KA sia
ly
_Ap
R6 a
Glu ophil phila a
s
se
il
o
o
s
ph
o
Dr
oumanat
ns
B_ A_Dr roso s
s _M u _R
ga
RII
I
n
Glu luRI IIC_Dlega Cele ns gan IK22_HIK2
G uR
e 5 _ g a l e GR I K R
R G
G l 6_C luR Cele _Ce
G
uR G a_ R3
Gl
4 lu
uR G
l
G
Grin2_L imulus
Grin2_Celegans
Gr in2 _A ply
sia
G rin 2_ Li
ng ul a
G ri n 2 _
Grin2
P
ri
a
p u lu s
_Lan
celet
G ri n
G r in
2A_C
2A_
io n a
Hum
Grin
G r in 2A_Ra a n
t
2
Gri A_Mo
u
n
se
Gr
in 2 A _
Gr 2 B _ Z e b r a
H
Gr i n 2 B _ u m a f i s h
in2
n
Mo
B_
Ra u s e
G Grin t
Gr rin2 2B_
C_
Ze
G in2
H
b
G rin2 C _ M u m r a f i s
rin C
a
h
2 C _R o u s n
e
G
a
rin G r _ Z e t
2D i n 2 b r
_R D _ a f i s
at Z e h
br
af
is
h
AMPA
0.4
Fig. 2.2. Bootstrap consensus phylogenetic tree. iGluR protein sequences for six
deuterostomes and nine protostomes with iGluR information available were aligned using
MUSCLE in MEGA7 and the tree was constructed using maximum likelihood and 1000
bootstrap replicates. iGluR subtypes are indicated
by different colors. Numbers indicate
0.4
bootstrap support and are displayed for boostrap values >0.6. The scale bar represents 0.4
substitutions per site. All species contain homologs for Grin1 and Grin2 subunits. An
ancestral AMPAR and kainate receptor gene was present in the common bilaterian
ancestor, but in the case of the kainate receptor, the monophyly is only weakly supported
by bootstrap values.
Fig2.3.ConservedmotifsinAplysiaandH.sapiens.BoxedregiondenotesSYTANLAAFLmotif
vitalforformationofthechannelporeandactivationgate,andinallsubunitsthishighselective
pressurehasresultedinhighlyconservedaminoacidsequencesoverlargeevolutionary
distance.Thismotifisalsoconservedinallothervertebratesinthisstudy.
44
45
"#
NMDA Binding
Glycine Binding
TM2
TM1
TM3
$#
%#
&#
Grin1&Aplysia
Grin1&Human
-./0&#12345/6#
%&'#
%%&#
%$%#
%"$#
%("#
%)(#
%*)#
%+*#
%,+#
$',#
$%'#
$$&#
$"%#
$($#
$)"#
$*(#
$+)#
$,*#
"'+#
"&,#
"$'#
""&#
"(%#
")$#
"*"#
"+(#
",)#
('*#
(&+#
(%,#
("'#
((&#
()%#
(*$#
(+"#
(,(#
)')#
)&*#
)%+#
)$,#
)('#
))&#
)*%#
)+$#
),"#
*'(#
*&)#
*%*#
*$+#
*",#
*)'#
**&#
*+%#
*,$#
+'"#
+&(#
+%)#
+$*#
+"+#
+(,#
+*'#
++&#
+,%#
,'$#
,&"#
,%(#
,$)#
,"*#
'#
-./0&#78960#
!&#
!%#
!$#
!"#
Fig. 2.4. Hydrophobicity plot of Grin1 in H. sapiens and Aplysia. Kite/Doolittle scale
was used to determine the hydrophobicity of each amino acid for the last 450 amino
acids, containing the ligand binding, transmembrane (TM), and intracellular C-terminal
domains. Despite large evolutionary distance between Aplysia and H. sapiens, only
substitutions with similar hydrophobicity are tolerated in TMs to maintain proper folding
of the ion channel. Predicted NMDA and glycine binding sites show many more
substitutions that result in changes in hydrophobicity.
46
A
Calculated number of transcripts of
iGluR genes
14000
AMPAR
NMDAR
12000
Kainate
Orphans
10000
*
8000
6000
4000
2000
0
Pedal
Pleural
B
Cerebral Abdominal
Ganglion
AMPA
NMDA
Buccal
Kainate/orphans
GluR1
GluR2
GluR6
Grin1-1
GluR3
GluR7
Grin1-2
GluR4
KA1
Grin2
GluR5
KA2
GluR8/10
Fig. 2.5. iGluR expression levels in Aplysia ganglia. The expression levels of each
subtype of iGluR per 100 ng of RNA were determined by quantitative real-time PCR.
Absolute copy number was calculated using standard curves. A) Comparisons of total
expression of each iGluR subtype in different Aplysia ganglia, with calculated number of
transcripts on the y-axis. Total kainate receptor subunit expression was significantly
greater than other subtypes in the pleural ganglia (one-way ANOVA, Tukey’s post-hoc,
p£0.05). B) Pie charts showing each subunit’s contribution to the total expression of its
subtype.
Chapter 3:
Altered expression of ionotropic L-Glutamate receptors in aged sensory neurons of
Aplysia
3.1 Summary
Changes in mRNA expression of ionotropic glutamate receptors (iGluR) were
studied in two cohorts of Aplysia at both sexual maturity and advanced age. Behavioral
aging of both cohorts was confirmed via the righting and tail withdrawal reflexes.
Previous studies have shown reduced behavioral performance in aged Aplysia is
correlated with reduced L-Glu current densities in buccal S cluster sensory neurons, as
well as in pleural ventral caudal sensory neurons that are involved in tail withdrawal.
Aging in both sets of sensory neurons resulted in significantly decreased expression of
iGluR subunits, particularly of the NMDA receptor subtype. This implies that age-related
declines in L-Glu responsiveness may be linked with reduced expression of receptors that
bind L-Glu. The NMDA receptor subtype was also expressed at significantly greater
levels than other iGluR subtypes. Thus, NMDA receptors may be a major contributor to
L-Glu mediated responses in Aplysia sensory neurons.
3.2 Background
The relatively simple nervous system of the model organism Aplysia californica
(Aplysia) allows for detailed studies of neuronal circuits that underlie many behaviors
(Rattan and Peretz 1981; Walters, et al. 1983a, b; Kandel 2001). These simple neuronal
circuits have been used to study changes in neuronal physiology that accompany aging in
47
48
the Aplysia nervous system (Hirsch and Peretz 1984; Fieber, et al. 2010; Kempsell and
Fieber 2014, 2015b). There is evidence that many neuronal circuits in Aplysia use Lglutamate (L-Glu) as their primary excitatory neurotransmitter (Dale and Kandel 1993;
Ha, et al. 2006). In this Chapter, previously described changes in L-Glu physiology in the
aging Aplysia nervous system were correlated with altered mRNA expression of L-Glu
receptors.
L-Glu binding to its constituent receptors on the postsynaptic membrane activates
both metabotropic and ionotropic L-Glu receptors. Metabotropic L-Glu receptors
(mGluR) act through activation of G-protein coupled second messenger cascades and can
induce changes in neuronal excitability and alterations in synaptic transmission
(Niswender and Conn 2010). Ionotropic L-Glu receptors (iGluR) mediate the majority of
fast synaptic transmission and regulate many processes in the central nervous system
(Dingledine, et al. 1999; Traynelis, et al. 2010a). iGluR are classified into three subtypes
based on selective agonists in vertebrates: N-methyl-D-aspartate (NMDA), α-Amino-3hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), and kainate. Functional iGluR are
formed as heterotetramers, consisting of iGluR subunits belonging to the same subtype.
Each iGluR subtype has distinct electrophysiological characteristics, and there is
evidence that analogues of each of these vertebrate subtypes exists in Aplysia, as
described below. The composition of Aplysia iGluR tetramers, if indeed its functional
iGluR are formed as in vertebrates, is unknown.
In vertebrates, NMDA iGluR (NMDAR) have relatively slow activation and
deactivation kinetics and are highly permeable to Ca2+ (Mayer and Westbrook 1987).
49
Activation of NMDAR and the opening of Ca2+ channels activates second messenger
cascades that have been implicated in the induction of neuronal plasticity (Riedel, et al.
2003). NMDAR are the most well-conserved iGluR subtype across bilaterian species,
including Aplysia (Chapter 2). This suggested that function of NMDAR may be
conserved between Aplysia and other bilaterian clades, including vertebrates (Chapter 2).
Application of NMDA to Aplysia buccal sensory neurons (BSC) in culture elicits small
amplitude currents and implies that NMDAR-like receptors are physiologically relevant
in Aplysia (Carlson and Fieber 2012; Fieber, et al. 2013).
The AMPA subtype of L-Glu receptors (AMPAR) mediate the majority of fast
excitatory transmission in the nervous system of vertebrates and have fast activation and
deactivation kinetics compared to NMDAR. AMPA application did not induce currents in
Aplysia BSC neurons in culture or in motoneurons of the abdominal ganglion (Trudeau
and Castellucci 1993; Carlson and Fieber 2012). However, the AMPAR antagonist 6cyano-7-nitroquinoxaline-2,3-dione (CNQX)-induced block of synaptic transmission,
suggesting that AMPA-like receptors exist in Aplysia (Trudeau and Castellucci 1993).
An ancestral AMPAR subtype gene was present in the bilaterian ancestor (Chapter 2), but
the Aplysia lineage has been subjected to several rounds of gene duplications. This
suggested that AMPAR may exist in Aplysia but have diverged to the point where
AMPA no longer binds as an agonist.
Kainate iGluR are present in both pre and post-synaptic terminal in vertebrates,
however their role is less clear than NMDAR and AMPAR. They are the most
evolutionarily recent iGluR subtype, and unlike NMDAR and AMPAR, vertebrate and
invertebrate kainate subunits do not form a clear monophyletic group (Chapter 2).
50
Kainate analogs elicited depolarizations and non-desensitizing inward currents in
glutamatergic BSC neurons of Aplysia, similar to L-Glu-induced effects in these neurons
(Wang, et al. 2013). Kainate is also an effective agonist in abdominal ganglion neurons
(Trudeau and Castellucci 1993). Thus, although they may not be well-conserved, kainatelike receptors may also execute physiologically relevant processes in Aplysia.
In Chapter 2, Aplysia iGluR subunits were placed into subtypes based on their
phylogenetic relationships with vertebrate iGluR subunits (See Table 2.2, in Chapter 2).
Grin1 and Grin2 subunits in Aplysia formed orthologous groups with vertebrate NMDAR
subunits, providing strong support for these subunits as NMDAR-like. Aplysia GluR1GluR5 and GluR8 formed a monophyletic clade with vertebrate AMPAR subunits, and
thus are predicted to play AMPAR-like roles. Kainate subunits formed only a weakly
supported monophyletic clade, with GluR7 and KA2 included in this clade. Aplysia KA1
and GluR6 were identified as orphan subunits with no clear relationship to vertebrate
iGluR subtypes.
Several behavioral responses in Aplysia decline incrementally as the animal ages.
Reduced function of iGluR in the underlying neuronal circuits has been implicated as a
contributor to these behavioral deficits. An example is the tail-withdrawal reflex (TWR).
During TWR, a stimulus applied to the tail causes withdrawal of the tail as a defensive
mechanism. Time to withdraw the tail increased significantly in animals of advanced age
(Kempsell and Fieber 2014). Pleural ventral caudal (PVC) sensory neurons in the pleural
ganglion make strong connections to tail motoneurons in the pedal ganglion and together
they comprise the neural circuit controlling TWR (Walters, et al. 1983a). PVC neurons
are highly glutamatergic, with 52% of PVC neurons responding to L-Glu application in
51
young animals (Kempsell and Fieber 2014). In aged animals with reduced TWR
behaviors, PVC neurons showed reduced L-Glu current density and the number of PVC
neurons responding to L-Glu application significantly decreased (Kempsell and Fieber
2014). Thus, reduced L-Glu responses in PVC sensory neurons were correlated with
reduced TWR and are believed a hallmark of aging in this neural circuit.
The biting response also exhibited significant behavioral deficits with age. Biting
amplitude decreased significantly in aged animals, while latency increased (Kempsell and
Fieber 2014). BSC sensory neurons innervating the biting response showed reduced
frequency of L-Glu responses and declines in L-Glu current densities, similar to those
observed in TWR (Kempsell and Fieber 2014). Together, these experiments implied that
aging of TWR and the biting response are due, at least in part, to reduced iGluR function
of their sensory neurons.
This study was an examination of iGluR in Aplysia sensory neurons as well as a
correlation of reduced L-Glu current density in PVC and BSC sensory neurons of aged
animals with altered transcription of iGluR subunits. We addressed which iGluR subunits
are most highly expressed in sensory neurons and may be involved in L-Glu mediated
responses. Also addressed was whether changes in transcription underlie reduced iGluR
current amplitude in aged animals. Changes in transcription can be used as a proxy for
the number of receptors expressed (Vogel and Marcotte 2012), and changes in iGluR
expression may be involved in reduced function of iGluR in aged Aplysia. iGluR
transcription in the abdominal ganglion, a heterogeneous ganglion containing both
sensory and motoneurons, was compared to iGluR transcription in the highly
glutamatergic clusters of PVC and BSC sensory neurons.
52
3.3 Materials and Methods
Aplysia from the University of Miami National Resource for Aplysia were reared
from egg masses of wild-caught animals. Two cohorts were reared and maintained
throughout their lifespan at a temperature of 13-15°C. Once the animals reached age 4
months they were held at a maximum of 5 per cage to allow for sufficient room to grow
large enough to reach sexual maturity. Animals were fed an ad lib diet of Agardhiella
subulata as previously described (Gerdes and Fieber 2006) throughout their lifespan.
Reflex completion times were measured in animals at sexual maturity (M; age 8.5
months) and advanced age (age 12 months) to quantify behavioral changes with age.
Sexual maturity for each cohort was designated as the first day an egg mass was laid.
Time to complete TWR and righting was recorded in 8 animals from each cohort, at each
stage of age. Mean reflex completion times were compared to Kempsell and Fieber
(2014) and determined that age 12 month animals in these two cohorts were Aged II
(AII).
Righting
Righting was measured at M and then again at AII in 16 animals (8 from each
cohort). Animals were placed in individual 48x27x20 cm cages with flowing seawater
and acclimated for five minutes before measurements of righting. To record righting, the
animal was picked up from the bottom of the cage and released upside down from the top
of the water column. Time was recorded from the when the animal was released until it
was able to right itself, adhere to the bottom of the cage, and take its first crawling step.
The measurement was recorded 3 times per individual with a minimum of 5 minutes
between trials, and the times averaged. Data was combined for all animals at each time
53
point and a Student’s t-test was used to determine significant differences in righting
reflex between age groups.
TWR
TWR was measured at both time points in the same individuals from both cohorts
that were analyzed for the righting reflex (n=8 for each cohort). After righting reflex
trials animals were allowed to rest for 5 minutes before the start of TWR measurements.
To record TWR the animal was first placed on its foot in the center of the cage, allowed
to adhere to the bottom, and was lightly held in place to measure initial resting length.
TWR was initiated by pressing a blunted 21-gauge needle to the tip of the animal’s tail
for approximately 1 sec to initiate withdrawal of the tail. To measure TWR, time was
recorded from the initiation of the reflex until the animal began to relax the tail after
withdrawal. Final retracted length was measured with a ruler to calculate percent tail
withdrawn. Measurements of duration and amplitude of TWR were recorded 3 times with
a minimum of 5 minutes between trials and the measurements were averaged. For
analysis, a Student’s t-test was used to calculate differences in TWR time and fraction of
body length withdrawn between age groups.
Weight
Weight measurements were recorded monthly for a random selection of 15
animals from each cohort beginning at age 3 months until the end of life, approximately
age 12.5 months in both cohorts. Excess water was drained from the parapodial cavity
prior to weight measurements.
54
RNA extraction of select ganglia for qPCR
RNA was extracted from six animals from each cohort at M and six animals from
each cohort at AII. The animals chosen were a randomly selected subset of the animals
tested for TWR and the righting at each age. Animals were first anesthetized by injection
of 1/6th body weight (by volume) of isotonic MgCl2. The CNS was removed by severing
the connectives and removal of all nervous system ganglia. Ganglia were immediately
rinsed in artificial sea water (ASW: 417mM NaCl, 10mM KCl, 10mM CaCl2, 55mM
MgCl2, and 15mM HEPES-NaOH, pH 7.6) after removal.
PVC neurons were separated from the remainder of the pleural ganglion by
pinning the ganglion tightly in a Sylgarded dish filled with ASW, as follows. Connective
tissue surrounding the ganglion was removed, and the sensory neuron clusters were
removed with a pair of fine forceps. Neuron clusters were immediately place in Qiagen
lysis buffer and both left and right PVC hemiganglia were pooled into a single sample.
The same procedure was used to remove BSC neurons from the buccal ganglion. Total
RNA from PVC and BSC neurons were extracted with the Qiagen RNeasy Micro Kit
following the manufacturer’s instructions. Genomic DNA contamination was removed
during RNA extraction according as manufacturer’s instructions (Qiagen).
The abdominal ganglia were rinsed with ASW and immediately placed in Trizol.
The tissue was ground in a bead homogenizer to break cells out of the connective sheath
prior to extraction RNA extraction. Total RNA was then extracted using Trizol Plus RNA
Purification kit following the manufacturer’s instructions (ThermoFisher).
Quantity of RNA from all tissues was analyzed using a NanoDrop (Model ND1000). One hundred ng of total RNA from each sample was reversed transcribed into
55
cDNA using SuperScript III reverse transcriptase (Invitrogen) with random hexamer
primers following the manufacturer’s suggestions. The resulting cDNA was diluted 1:5
with nuclease free water to provide a final concentration of 1 ng RNA/1 µl cDNA for
qPCR.
qPCR
qPCR reactions were executed on a Stratagene Mx3005P with SYBR Green
master mix (Invitrogen). Each reaction consisted of 10 µl of SYBR Green master mix, 2
µl of cDNA and 0.2 µM of forward and reverse primers in a total volume of 20 µl.
Thermal cycler conditions were as follows: 95°C for 10 minutes, followed by 40 cycles
of 95°C for 15 sec, 58°C for 30 sec, 72°C for 30 sec. All 12 iGluR subunits and the splice
variant of Grin1 identified were tested (see Table 2.2, Chapter 2 for list of iGluR genes).
Primer sequences and amplicon lengths for each gene have been previously described
(see Table 2.1, Chapter 2).
For cohort 1, these ganglia and sensory neuron clusters from 6 animals at M and
AII were analyzed with qPCR to quantify changes in iGluR expression in the nervous
system. Two technical replicates were performed for each biological replicate.
Absolute copy number for each replicate was calculated using serial dilutions of a
known quantity of plasmid containing the target amplicon. PCR products for each gene
were gel purified with a Qiagen gel purification kit and cloned into electrocompetent E
coli cells using a TOPO TA cloning kit (ThermoFisher). Transformed bacterial colonies
were plated onto LB agar plates with ampicillin and blue/white screened with X-gal to
identify colonies containing the desired plasmid. Five bacterial colonies were picked
from each plate and bacteria grown in a shaker overnight at 37°C. Plasmids were then
56
purified with a Qiagen mini-prep plasmid kit and quantified on a Nanodrop to determine
plasmid concentration. Serial dilutions from 106-100 plasmid copies were run in triplicate
on each plate to calculate a standard curve for absolute copy number calculation.
For cohort 2, PVC and BSC neuron clusters from 4 animals were tested at each
time point. These samples were run in technical triplicate instead of duplicates as were
run for cohort 1. In this cohort we tested the expression of several of the most highly
expressed iGluR subunits: Grin1-1, Grin1-2, Grin2, GluR1, GluR5, and GluR7.
Instead of absolute quantification using a standard curve on each plate,
quantification of target transcripts was normalized to GAPDH using the Gene
Expression’s CT Difference (GED) formula (Schefe, et al. 2006). Briefly, this calculation
is a modified version of the DDCT method previously described (Pfaffl 2001). The DDCT
method relies on calculating the efficiency of the reaction based on the slope of one cycle
in the exponential phase of the reaction. It has been shown that the efficiency of the
qPCR reaction may vary in different parts of the qPCR plate and varying salt
concentrations in different samples (Schefe, et al. 2006). The GED formula accounts for
different efficiencies in each reaction by using calculating the efficiency of each qPCR
reaction independently using Real-time PCR miner (Zhao and Fernald 2005). These
efficiencies were then used to calculate relative transcription for each sample as
previously described (Pfaffl 2001; Schefe, et al. 2006).
57
3.4 Results
Behavioral aging in two cohorts
Average weight declined beginning approximately two months prior to AII
measurements in both cohorts, consistent with previously studies in aged Aplysia of
reduced mass during advanced stages of aging (Fig. 3.1A). Time to right significantly
increased in cohort 1 (Fig. 3.1B, 13.3±0.39 sec (mean±standard error) at M, 22.0±0.59
sec at age 12 months, p£0.05, Student’s t-test). TWR also increased significantly in AII
animals compared to M in cohort 1 (Fig. 3.1C, 13.7±0.58 sec at M, 24.2 ± 0.58 sec at age
12 months, p£0.05, Student’s t-test). Significantly increased righting and TWR times in
AII were also observed in cohort 2 (Fig. 3.1B and C; Righting: 13.3±0.47 sec at M,
23.7±0.68 sec age 12 months; TWR: 13.6±0.33 sec at M, 27.2±0.37 sec at age 12
months, p£0.05, Student’s t-test). The observed TWR ³20.1 sec and righting ³18.1 sec
for age 12 month animals of both cohorts corresponded to AII based on published stages
of aging in Aplysia (Kempsell and Fieber 2014).
The amplitude of TWR, measured as the fraction of body length withdrawn in
response to tail touch, decreased significantly in AII animals in both cohorts (Fig. 3.1D;
cohort 1: 26.1±1.2% at M, 13.7±0.7% at AII; cohort 2: 27.5±2.0% at M, 15.8±1.0% at
AII, p£0.05, one-way ANOVA on a logistic-regression with quasibinomial distribution).
Thus, reflex times increased in AII while the amplitude of TWR decreased.
The most highly expressed iGluR subunits in three nervous system tissues
Differential expression of iGluR subunits was first studied in abdominal and
sensory neuron clusters separately, without accounting for differences in age. These data
were analyzed using data from cohort 1 because calculation of absolute copy number in
58
this cohort allowed for direct comparisons of genes from different qPCR plates. Grin1-1
showed significantly greater expression than all other iGluR subunits in each of the three
tissues tested (Fig. 3.2A-C, 30,779±4,301 copies (mean±standard error) in PVC neurons,
27,095±3,572 copies in BSC neurons, 20,646±2,504 copies in abdominal ganglia,
p£0.05, one-way ANOVA, Tukey’s post-hoc test).
In PVC sensory neurons, GluR1 was the next most highly expressed subunit after
Grin1-1, and GluR1 showed significantly greater expression than all less-expressed
subunits except GluR7. GluR7 in PVC neurons was expressed at significantly greater
copies than only GluR2 and KA2 (Fig. 3.2A). In BSC sensory neurons, GluR1 and
Grin1-2 were expressed at significantly higher copies than all other subunits after Grin1-1
(Fig. 3.2B). GluR7, GluR1, and Grin1-2 subunits were expressed significantly higher in
the abdominal ganglia than were other iGluR subunits, except Grin1-1 (Fig. 3.2C).
Differential expression of iGluR subunits in different parts of the nervous system
Next, each subunit was compared across the ganglia to reveal patterns of
differential subunit expression in the different ganglia. This analysis was calculated for
cohort 1 only, and expression from M and AII animals was pooled. Apparent once more
was the dominant expression in all 3 tissues of Grin1-1, expressed at a minimum of >2
times the expression of other subunits in all tissues (Fig. 3.3, p£0.05 for each tissue, oneway ANOVA, Tukey’s post-hoc). Several differences in expression of other, individual
subunits across tissues, however, were clear. GluR5 was expressed at significantly greater
copies in PVC neurons than in either BSC neurons or the abdominal ganglia (Fig. 3.3,
4,015±643 copies (mean±standard error) in PVC neurons, 386±73 copies in BSC
neurons, 1282±194 copies in abdominal ganglia, p£0.05, one-way ANOVA, Tukey’s
59
post-hoc). Grin1-2, GluR2, and GluR4 were expressed significantly more in BSC sensory
neurons than the other tissues (Grin1-2: 5,170±719 copies in PVC neurons, 11,263±1350
copies in BSC neurons, 5,269±536 copies in abdominal ganglia; GluR2: 349±90 copies
in PVC neurons, 864±185 copies in BSC neurons, 400±108 copies in abdominal ganglia;
GluR4: 739±87 copies in PVC neurons, 2,253±265 copies in BSC neurons, 385±33
copies in abdominal ganglia, p£0.05 for each gene, one-way ANOVA, Tukey’s posthoc). Grin2 was expressed at significantly lower levels in BSC neurons than in any other
ganglia. (2,970±264 copies in PVC neurons, 1,407±144 copies in BSC neurons,
3490±342 copies in abdominal ganglia, p£0.05, one-way ANOVA with Tukey’s posthoc) In the abdominal ganglia, GluR7 expression was significantly greater than in the
other tissues, while KA1 and KA2 expression was lower in the abdominal ganglia than in
other tissues (GluR7: 6,772±753 copies in PVC neurons, 6,258±574 copies in BSC
neurons, 8,741±770 copies in abdominal ganglia; KA1: 1,559±199 copies in PVC
neurons, 2,036±120 copies in BSC neurons, 1,036±96 copies in abdominal ganglia; KA2:
541±45 copies in PVC neurons, 451±51 copies in BSC neurons, 368±26 copies in
abdominal ganglia, p£0.05 for each gene, one-way ANOVA with Tukey’s post-hoc).
Expression of iGluR subunits with age
Decreased TWR in AII was previously correlated with reduced L-Glu current
density in PVC sensory neurons for the reflex (Kempsell and Fieber 2014). Whether
declines in TWR were correlated with altered expression of iGluR subunits was
investigated by comparing iGluR expression in PVC neurons of M and AII animals. In
cohort 1, AII PVC sensory neurons exhibited significantly decreased expression of
several iGluR subunits; these were both splice variants of Grin1: Grin1-1 and Grin1-2
60
(Fig. 3.4A; Grin1-1 AII:M expression 0.56±0.08, Grin1-2: AII:M expression 0.60±0.13,
p£0.05 for each gene, Student’s t-test). Grin2, the only remaining NMDAR subunit,
showed no change in AII compared to M. No changes in expression of any AMPAR
subunits in aged PVC neurons were found. KA1 and GluR6, orphan receptors without a
clear relationship to any iGluR subtype, had significantly decreased expression in AII
(Fig. 3.4A; KA1 AII:M expression 0.47±0.10; GluR2 AII:M expression 0.46±0.09,
p£0.05 for each gene, Student’s t-test).
Select subunits were analyzed in cohort 2 to attempt to substantiate differential
expression observed in cohort 1 (Fig 3.4B). Distinct differences were apparent in cohort 2
PVC neurons. Although Grin1-1 was significantly down-regulated in cohort 2 AII PVC
neurons (AII:M expression 0.38±0.04, p£0.05, Student’s t-test), as had been observed in
cohort 1, Grin1-2 expression was not significantly changed, representing a departure
from the difference between M and AII Grin1-2 in cohort 1. GluR1, an AMPAR subunit,
exhibited reduced expression in cohort 2 AII PVC neurons, but this subunit had not been
differentially expressed with age in cohort 1 (AII:M expression 0.41±0.03, p£0.05,
Student’s t-test). Transcript copy numbers per 100 ng of RNA were also calculated for
cohort 1, and age-related changes in iGluR expression described above are displayed as
copy numbers in Fig. 3.5.
BSC sensory neurons innervate the biting reflex and were characterized by
reduced L-Glu current density in AII animals (Kempsell and Fieber 2014). Expression of
iGluR genes in BSC neurons of cohorts 1 and 2 were compared (Fig. 3.6). Grin1-1 and
Grin1-2 were significantly downregulated in AII BSC sensory neurons of cohort 1, just as
in PVC neurons of this cohort (Fig. 3.6A, Grin1-1 AII:M expression 0.59±0.08, Grin1-2
61
AII:M expression 0.57±0.05, p£0.05 for each gene, Student’s t-test). These subunits were
also significantly down-regulated in cohort 2 (Fig. 3.6B, Grin1-1 AII:M expression
0.64±0.04, Grin1-2 AII:M expression 0.56± 0.06, p£0.05, Student’s t-test), whereas only
Grin1-1 had been significantly downregulated in cohort 2 PVC neurons. In addition to
Grin1 subunits, the AMPAR subunit GluR1 had significantly decreased expression in
cohort 2, just as in the AII PVC neurons of cohort 2 (Fig. 3.6B, AII:M expression
0.43±0.02, p£0.05, Student’s t-test). AMPAR subunit GluR3 and orphan subunit GluR6
had significantly reduced expression in cohort 1 AII BSC neurons (Fig. 3.6A, GluR3
AII:M expression 0.61±0.05, GluR6 AII:M expression 0.69±0.05, p£0.05 for each gene,
Student’s t-test). Transcript copy numbers per 100 ng of RNA were also calculated for
cohort 1, and age-related changes in iGluR expression described above for BSC neurons
are displayed as copy numbers in Fig. 3.7.
Cohort 1 AII abdominal ganglion are displayed as transcript copy numbers in Fig.
3.8. The abdominal ganglion showed no significant changes in expression of any iGluR
subunits (Fig. 3.8). The abdominal ganglion was not tested in cohort 2.
Subunit composition
Percent contribution of each subunit to the total expression for each subtype was
calculated to gain insights about possible components of complete functional iGluR in
Aplysia. These data were calculated using cohort 1 only, because calculation of absolute
copy number in this cohort allowed for comparison of expression data across different
qPCR runs. It has previously been shown that Aplysia NMDAR and AMPAR subunits
form monophyletic groups and thus may form complete receptors (Chapter 2). The
62
makeup of kainate-like receptors in Aplysia is less clear, therefore we were unable to
calculate subunit composition for this subtype.
In vertebrates, NMDAR are comprised of a heterotetramer of two Grin1 subunits
and two Grin2 subunits. Thus, it is to be expected that they are expressed in
approximately equal amounts. In both PVC and BSC sensory neurons, the two splice
variants of Grin1 together comprised >90% of total NMDAR subtype expression (Fig.
3.9A and B). This was true in M as well as AII PVC and BSC neurons. There were no
statistically significant changes in the percent contribution of any NMDAR subunit
between M and AII (Fig. 3.9A and B, p£0.05, one-way ANOVA on a logistical
regression). For example, Grin1-1 contributed ~80% of NMDAR expression in both M
and AII PVC neurons. Similarly, expression of the two Grin1 splice variants comprised
~85% of total NMDAR subunit expression in abdominal ganglia, and did not vary with
age (Fig. 3.9C). Thus, Grin1 contributed the majority of NMDAR mRNA expression in
the ganglia tissues studied, with NMDAR subunit percentages not significantly altered
during aging.
AMPAR subunits were analyzed similarly. GluR1 was the predominately
expressed subunit in PVC and BSC sensory neurons, representing ~60% of total AMPAR
expression in both sensory neuron clusters (Fig. 3.10A and B). GluR1 was also the most
highly expressed subunit in the abdominal ganglia at ~70% of total AMPAR expression
(Fig. 3.10C). This was a significantly higher percentage of total AMPAR expression than
observed in PVC and BSC neurons (p£0.05, one-way ANOVA on a logistical regression,
Tukey’s post-hoc). In all three tissues there were no significant differences in percent
composition of any AMPAR subunit in AII animals (Fig. 3.10A-C).
63
3.5 Discussion
In this study, mRNA expression of iGluR subunits in mature and aged sensory
neuron clusters of the Aplysia nervous system was quantified. Changes in mRNA
expression can be used as a proxy for decreased in protein synthesis that results in
phenotypic changes (Vogel and Marcotte 2012). Therefore, decreased expression of
iGluR subunits observed in aged sensory neurons in this study may directly influence the
amount of iGluR receptor protein, and decrease the number of receptors available for LGlu mediated transmission. There is also the possibility that transcriptional changes may
not be directly correlated with changes in protein, such as changes in turnover rate
requiring altered transcription to maintain the same function. However, there is ample
evidence to suggest that altered transcription of receptor genes is correlated with altered
physiology.
For example, in the plainfin midshipman, males produce a vocal hum as a
courting behavior only during the breeding season. Females with mature eggs exhibited a
strong phonotaxis to male humming frequencies that is only present during this time, and
is lost the remainder of the year (McKibben and Bass 1998; Sisneros and Bass 2003;
Zeddies, et al. 2010). Increased sensitivity to male vocal hums during the mating season
correlated with an increased number of calcium-activated potassium (BK) channels in
auditory hair cells (Coffin, et al. 2012). Expression levels of two isoforms of BK
channels quantified by qPCR, showed upregulation in reproductive compared to nonreproductive females (Rohmann, et al. 2013). Hence, increased channel expression during
mating season is correlated with increased transcription of BK channel genes.
64
Altered expression in euryhaline fish, such as the Mozambique tilapia, is
correlated with the drastic physiological changes necessary to maintain osmotic balance
in rapidly changing salinity regimes. The hormone prolactin plays an important
osmoregulatory role when euryhaline fish encounter freshwater, by preventing water
uptake and excessive loss of ions (Manzon 2002). Expression of prolactin receptors in the
gills of Mozambique tilapia quantified by qPCR were significantly increased when
exposed to freshwater, consistent with its required induction for freshwater adaptation
(Pierce, et al. 2007).
In human cognitive disorders, particularly schizophrenia, reduced NMDAR
function is a major component of the disease (Tsai and Coyle 2002; Paoletti and Neyton
2007). Reducing NMDAR expression in transgenic mice recapitulates the symptoms,
suggesting that reduced NMDAR expression leads to observed physiological dysfunction
involving NMDAR that is prevalent in the disease (Mohn, et al. 1999). Together these
studies show that reduced expression of constituent receptors is be correlated with altered
physiology during major life history events. Thus, qPCR results on iGluR in this study
can be expected to have physiological relevance in the Aplysia nervous system.
Grin1-1, a subunit of the NMDAR subtype, was expressed at greater than twofold more than any other iGluR subunit in PVC sensory neurons, BSC sensory neurons,
and the abdominal ganglion. Grin1 mRNA previously has been demonstrated to be
expressed throughout the Aplysia nervous system (Ha, et al. 2006), and is also widely
expressed in Lymnaea neurons and mammalian brains (Akazawa, et al. 1994; Goebel and
Poosch 1999; Ha, et al. 2006). Ubiquitous expression of Grin1-1 in the nervous system
suggests an important role for NMDAR in iGluR physiology across Bilateria. The greater
65
expression of Grin1-1 compared to other subunits suggests that iGluR in Aplysia is
mostly comprised of Grin 1-1, and that most L-Glu-mediated responses studied in
Aplysia may be Grin1-1-mediated.
Other iGluR subunits were expressed in significantly higher abundance in one
ganglion, and therefore may be more greatly involved in mediating iGluR transmission in
that ganglion. GluR5 was expressed at significantly greater copies in PVC neurons than
in other ganglia, and Grin1-2, GluR2, and GluR4 were all more highly expressed in BSC
neurons. Thus, different parts of the Aplysia nervous system may emphasize different
iGluR subunits in iGluR-mediated neurotransmission.
Not only were Grin1-1 subunits highly expressed in this study, they also exhibited
reduced expression in AII PVC and BSC neurons. If Grin1-1 plays the largest role is LGlu mediated responses, as the expression data indicate, its down-regulation would be
predicted to have an impact on L-Glu-induced current densities in aged animals. Fieber et
al. (2010) noted decreased density and frequency of iGluR-activated whole cell currents
in PVC neurons from Aplysia near the end of life, without changes in BSC neuron iGluR
currents. Later, Kempsell and Fieber (2014) demonstrated that AII PVC and BSC sensory
neurons both had reduced L-Glu-induced current density. This suggested that the reflex
declines noted in aging are correlated with either impaired iGluR function or iGluR
absence in the neurons that underlie the behaviors. Reduced Grin1-1 expression in AII
sensory neurons observed in this study is proposed as a contributor to deficits in L-Glumediated excitability that, in turn, contribute to the impaired TWR and biting reflex.
Declines in mRNA and protein expression of Grin1 subunits have been described
in the brain of aged rats, mice, and macaque monkeys, suggesting that Grin1 subunits are
66
highly susceptible to the effects of aging (Gazzaley, et al. 1996; Hof, et al. 2002; Shi, et
al. 2007; Newton, et al. 2008). Reduced NMDAR mRNA expression in aged mice
resulted in reduced protein expression that was greater than observed reductions in
mRNA (Magnusson, et al. 2002). Therefore, it is reasonable to propose that reduced
expression of NMDAR subunits observed here led to a decreased number of NMDAR
proteins and fewer iGluR in sensory neurons.
Significantly reduced expression of other subunits was also found, and may
contribute to impaired L-Glu-mediated reflex pathways and impaired sensory neuron
excitability in AII. KA1 and GluR6 were reduced in AII PVC neurons while GluR3 and
GluR6 were reduced in BSC neurons. No subunits had increased expression in AII. Agerelated reductions in iGluR expression have been described in several vertebrate species
(Wardas, et al. 1996; Hof, et al. 2002; Shi, et al. 2007), with reduced binding of L-Glu in
aged neurons linked to reduced numbers of functional iGluR (Wenk and Barnes 2000).
Thus reduced expression is likely a reasonable proxy for reduced glutamatergic
neurotransmission and reflex function in aging in these models.
The effect of aging on iGluR expression varied between PVC and BSC neurons,
and in the two different cohorts. Some expression differences, such as reduced Grin1-1
expression, were observed in both cohorts and both tissues. Other iGluR genes with
reduced expression varied between different cohorts and ganglia. For example, GluR1
was significantly down-regulated in AII of both sensory neurons clusters of cohort 2, yet
was not differentially expressed in cohort 1. The discrepancy in GluR1 expression in
PVC neurons between cohorts did not manifest as a change in the behavioral aging of
TWR. Different rates of aging of iGluR physiology also occurred in different cohorts and
67
ganglia of Aplysia. Fieber et al. (2010) found no iGluR current differences in BSC
neurons of three cohorts with age, while these same cohorts showed differences in iGluR
currents in PVC neurons. In contrast, Kempsell and Fieber (2014) did observe iGluR
current differences with age in BSC neurons of a different cohort. Thus, there are both
similarities and differences in molecular and physiological aging of L-Glu, depending on
the cohort and neuronal clusters examined. Physiological outcomes and behaviors may
not need to map 1:1 with differences in expression to contribute to declines in nervous
system function with age.
Other neurons in the Aplysia nervous system showed age-related physiological
deficits correlated with reduced receptor mRNA. The cholinergic R15 bursting neuron in
the abdominal ganglion that showed reduced responsiveness to acetylcholine in aged
animals was correlated with its decreased expression of acetylcholine receptors
(Akhmedov, et al. 2013) These results parallel those found in the sensory neurons in this
study, with reduced neurotransmitter actions in isolated neurons coinciding with
decreased expression of its receptors.
While several iGluR subunits decreased expression in AII PVC and BSC neurons,
there were zero changes in iGluR expression in the abdominal ganglion. This suggests
that iGluR transmission may not have been affected in the abdominal ganglia and that
aging of iGluR occurred differently in different parts of the nervous system. Specific
gene expression changes that vary between different individual neurons during aging has
also been described in other neurons of the Aplysia nervous system (Moroz and Kohn
2010; Kadakkuzha, et al. 2013). Moroz and Kohn (2010) found that in two cholinergic
neurons, R2 in the abdominal ganglia and LPl1 in the left pleural ganglion,
68
approximately half of the genes significantly affected by aging occurred in either R2 or
LPl1, but not in both. For example, potassium channel subunits increased expression in
aged R2 neurons, but reduced expression in aged LPl1 neurons. Inhibitor of apoptosis
homolog protein was highly expressed in aged R2 and was relatively absent in LPl1
(Moroz and Kohn 2010). Transcription factors CREB1 and CREB2 also show highly
variable patterns of expression in aging and showed increased, decreased, or no changes
in expression in different individual neurons of the abdominal ganglia (Kadakkuzha, et al.
2013). Hence, differential aging of different neurons is common in Aplysia, and we have
extended these findings to include differential aging of iGluR subunits.
Differential aging of iGluR also occurs in the vertebrate brain. Reduced density of
NMDAR-mediated currents during aging has been reported in many regions of the
vertebrate brain (Magnusson and Cotman 1993; Nicolle, et al. 1996; Wardas, et al. 1996;
Mitchell and Anderson 1998). In those studies, receptor density decreased 20-50% in
aged animals. It has also been reported that in other regions of the vertebrate brain,
NMDAR subunits may be upregulated or experience no changes in expression during
aging (Jouvenceau, et al. 1998; Potier, et al. 2000).
It would be expected that complete NMDAR in Aplysia are formed from two
Grin1 and two Grin2 subunits in Aplysia, based on vertebrate NMDAR that consist of
obligatory heterotetramers, formed from two Grin1 subunits and either two Grin2
subunits or two Grin3 subunits (Traynelis, et al. 2010a; Paoletti 2011). NMDAR are the
most well-conserved iGluR subtype across bilaterian species, including Aplysia (Chapter
2), and it is reasonable to assume that functional Aplysia iGluR are formed from similar
heterotetramers. Grin3 subunits have not been characterized in Aplysia and many other
69
protostomes (Chapter 2). Grin1 was expressed ~10 times higher than Grin2 in all three
Aplysia ganglia. If Aplysia NMDAR are formed as heterotetramers of Grin1 and Grin 2,
it may be that Grin1 turnover rates are much higher than Grin2, requiring greater
transcription to maintain higher turnover rates. Another possibility is that Grin1 may
form homotetramers in Aplysia. However, expression of Aplysia Grin1 only in Xenopus
oocytes did not induce L-Glu currents (Ha, et al. 2006), casting doubt on the hypothesis
that high expression of Grin1-1 may indicate homotetramer formation. These results
suggest that if Aplysia NMDAR are tetrameric, the makeup of complete receptors may
not be able to be predicted based on expression levels.
Altered composition of the subunits comprising complete tetrameric AMPAR and
NMDAR is associated with decreased L-Glu-mediated transmission in the aging
vertebrate brain. Variations in the subunits that make up a complete receptor alter the
biophysical, pharmacological, and signaling properties of the receptor and can confer
different messages to the same stimulus (Paoletti 2011; Paoletti, et al. 2013). No changes
in the percent contribution of any subunit were found during aging for NMDAR or
AMPAR in PVC and BSC sensory neurons. If complete tetrameric receptors in Aplysia
are formed from subunits within each subtype as predicted, these results suggest that
reduced L-Glu responsiveness of aged PVC and BSC neurons is likely due directly to
reduced expression of iGluR subunits, rather than altered receptor composition.
In conclusion, quantification of mRNA can be used as a powerful tool to examine
underlying changes in the nervous system. During major life history events, such as
aging, changes in mRNA expression in the nervous system can be correlated with and
can be a proxy for alterations in neuronal physiology. To follow up on this idea, Chapter
70
4 examines whole-transcriptome mRNA in M and AII PVC neurons for changes that may
be correlated with overall reduced neuronal activity of aged sensory neurons.
71
Cohort 1
Cohort 2
30
400
300
200
100
0
2
Time to relax tail (s)
C
4
20
15
10
5
*
25
*
20
15
10
5
0
08
6
10 12
8.5 Age (mos)
Age (mos)
30
*
25
20
15
10
5
0
8.5
12
Age (mos)
*
D
Percent tail withdrawn
Weight (g)
500
30
B
25
Time to right (s)
Time to relax tail (s)
600
A
12
35
30
25
20
15
10
5
0
8.5
12
Age (mos)
*
*
8.5
12
Age (mos)
Fig 3.1. Behavioral and weight correlates of aging in two cohorts of Aplysia. All error
bars represent ±SE of the mean. A) Animal mass estimated from wet weight decreased
beginning approximately 2 months before AII measurements (n=15). Dotted lines
represent time points animals were sacrificed for qPCR. B) Mean time to right increased
significantly in AII. C) TWR increased significantly in AII animals compared to M
animals in both cohorts. D) Amplitude of tail withdrawal as fraction of the initial length
withdrawn decreased significantly in AII.
* denotes significantly different than M (p£0.05, Student’s t-test)
72
A
B
BSC
Abdominal
GluR1 Grin1-1 Grin1-2 GluR7
GluR1 Grin1-1 Grin1-2 GluR7
PVC
GluR1
Grin1-2
Grin2
GluR2
GluR3
GluR4
GluR5
GluR6
GluR8
KA1
KA2
Grin1-1
GluR7
Grin1-2 GluR2
Grin2
KA2
GluR1
GluR2
GluR3
GluR4
GluR5
GluR6
GluR7
GluR8
KA1
KA2
C
Grin2
GluR2
GluR3
GluR4
GluR5
GluR6
GluR8
KA1
KA2
Grin1-2
Grin2
GluR1
GluR2
GluR3
GluR4
GluR5
GluR6
GluR7
GluR8
KA1
KA2
Grin2 GluR5
GluR2 GluR8
GluR3 KA2
GluR4
GluR5
GluR6
GluR8
KA1
KA2
Grin1-2
Grin2
GluR2
GluR3
GluR4
GluR5
GluR6
GluR8
KA1
KA2
Grin1-2
Grin2
GluR1
GluR2
GluR3
GluR4
GluR5
GluR6
GluR7
GluR8
KA1
KA2
Grin2
GluR2
GluR3
GluR4
GluR5
GluR6
GluR8
KA1
KA2
Grin2
GluR2
GluR3
GluR4
GluR5
GluR6
GluR8
KA1
KA2
Fig. 3.2. Differential expression of iGluR subunits in sensory neuronal clusters and
abdominal ganglia of both Mature and Aged II animals. Blue circles represent
subunits that were expressed at significantly greater copies than all subunits listed
underneath them in A) PVC sensory neurons B) BSC sensory neurons and C) whole
abdominal ganglia. Grin1-1 was the most highly expressed subunit in all three ganglia
tested and was expressed at significantly greater copies than any other iGluR subunits,
including other significantly differentially expressed subunits (all other blue circled
subunits; p£0.05, one-way ANOVA, Tukey’s post-hoc).
*Subunits listed underneath are ordered by subtype (NMDAR, AMPAR,
kainate/orphans)
73
PVC
35000
BSC
Abdominal
30000
Calculated number of transcripts
25000
20000
15000
*
*
10000
*
5000
*
*
*
*
*
*
NMDA
AMPA
GluR6
KA1
KA2
GluR7
GluR8
GluR5
GluR4
GluR3
GluR2
GluR1
Grin2
Grin1-2
Grin1-1
0
Kainate and orphans
Subunit
Fig. 3.3. Calculated number of iGluR transcripts in different ganglia. Many iGluR
subunits were significantly differentially expressed across different ganglia. Number of
transcripts is calculated per 100 ng of RNA. This data does not account for age-related
changes in expression and includes both M and AII animals.
*denotes significantly differentially expressed compared to all other ganglia (p£0.05,
one-way ANOVA, Tukey’s post-hoc test)
74
A
Relative expression compared to
Mature
2.5
2
1.5
1
*
*
*
*
0.5
0
AMPA
NMDA
Kainate/orphans
Subunit
B
Relative expression compared to Mature
1.4
1.2
1
0.8
0.6
*
*
0.4
0.2
0
AMPA
NMDA
Kainate
Subunit
Fig. 3.4. Relative expression of iGluR subunits in PVC sensory neurons of Mature
and Aged II animals from two cohorts. A) Relative expression for all iGluR subunits in
AII compared to M (dotted line) in PVC sensory neurons of cohort 1. Several NMDAR
and kainate/orphan receptor subunits were significantly down-regulated in AII PVC
neurons. There were no changes in expression to any AMPAR subunits. B) Relative
expression of iGluR subunits in PVC sensory neurons of AII animals compared to M in
cohort 2. Only select subunits with the greatest expression in cohort 1 were analyzed in
cohort 2. Grin1-1 and GluR1 had significantly reduced expression in AII compared to M.
AII expression level for each gene was normalized against expression in M PVC neurons
(dotted line).
*denotes significantly reduced expression compared to Mature (Student’s t-test, p£0.05;
n=6 for cohort 1, n=4 for cohort 2)
75
Mature
Calculated number of transcripts
50000
Aged II
40000
30000
*
20000
10000
*
*
KA1
GluR6
*
NMDA
AMPA
KA2
GluR7
GluR8
GluR5
GluR4
GluR3
GluR2
GluR1
Grin2
Grin1-2
Grin1-1
0
Kainate and orphans
Subunit
Fig. 3.5. Absolute transcript quantification of iGluR subunits in PVC sensory
neurons of Mature and Aged II animals from cohort 1. Expression changes with age
of iGluR genes from cohort 1 displayed in Fig. 3.4, shown as the number of transcripts
per 100 ng of RNA.
*denotes significantly reduced expression compared to Mature
76
A
Relative expression compared to
Mature
1.2
1
0.8
*
*
*
*
0.6
0.4
0.2
0
AMPA
NMDA
Kainate/orphans
Subunit
B
Relative expression compared to Mature
1
0.9
0.8
0.7
*
*
0.6
0.5
*
0.4
0.3
0.2
0.1
0
AMPA
NMDA
Kainate
Subunit
Fig. 3.6. Relative expression of iGluR subunits in BSC sensory neurons of Mature
and Aged II animals from two cohorts. A) Relative expression for all iGluR genes in
AII BS sensory neurons compared to M in cohort 1. Grin1-1 and Grin1-2 NMDAR
subtype genes were significantly down-regulated in AII. GluR3 and GluR6 also had
significantly reduced expression in AII BSC neurons. B) Relative expression of iGluR
subunits in AII BSC sensory neurons in cohort 2. Grin1-1 and Grin-2 were significantly
down-regulated, which was also observed in cohort 1. GluR1 was also down-regulated in
AII. AII expression levels were normalized against expression in M PVC neurons (dotted
line).
*denotes significantly reduced expression compared to M (Student’s t-test, p£0.05; n=6
for cohort 1, n=4 for cohort 2)
77
50000
Mature
Aged II
45000
40000
Calculated number of transcripts
35000
30000
25000
*
20000
15000
*
10000
5000
*
*
GluR6
KA1
KA2
GluR7
GluR8
GluR5
GluR4
GluR3
GluR2
GluR1
Grin2
Grin1-2
Grin1-1
0
Fig 3.7. Absolute transcript quantification of iGluR subunits in BSC sensory
neurons of Mature and Aged II animals from cohort 1. Expression changes with age
of iGluR genes from cohort 1 displayed in Fig. 3.6, shown as the number of transcripts
per 100 ng of RNA.
*denotes significantly reduced expression compared to Mature
78
50000
Mature
Aged II
45000
40000
Calculated number of transcripts
35000
30000
25000
20000
15000
10000
5000
NMDA
AMPA
GluR6
KA1
KA2
GluR7
GluR8
GluR5
GluR4
GluR3
GluR2
GluR1
Grin2
Grin1-2
Grin1-1
0
Kainate and orphans
Subunit
Fig. 3.8. Calculauted number of iGluR transcripts in Mature and Aged II whole
abdominal ganglion of cohort 1. In the abdominal ganglion, zero iGluR subunits
exhibited altered expression in AII animals.
79
Mature
2000
B
1800
Calculated number of transcripts
% of total NMDAR expression
100 1600
90 1400
80
70
60
1200
1000
800
50
40
30
20
10
600
400
200
0
Grin1
Grin2
Subunit
0
100
% of total NMDAR expression
A
Aged II
90
80
70
60
50
40
30
20
10
0
Grin1-1
C
Grin1-2
Subunit
Grin2
Grin1-1
Grin1-2
Subunit
Grin2
% of total NMDAR expression
100
90
80
70
60
50
40
30
20
10
0
Grin1-1
Grin1-2
Subunit
Grin2
Fig. 3.9. Percent composition of NMDAR subtype subunits. A) Expression of
NMDAR subunits in PVC sensory neurons showed no significant change in percent
composition during aging. B) Percent composition also did not change significantly with
age in BSC sensory neurons. C) Composition of NMDAR was not significantly altered in
abdominal ganglia of AII animals.
p>0.05 for all comparisons
80
Mature
2000
Aged II
1800
30
600
400
20
Subunit
Grin2
GluR8
GluR5
GluR3
Grin1
GluR4
0
GluR1
0
GluR2
10
200
Subunit
C
60
50
40
30
20
10
0
GluR8
40
800
70
GluR5
50
1000
80
GluR4
60
1200
90
GluR3
70
1400
GluR2
80
100
% of total AMPAR expression
90
Calculated number of transcripts
% of total AMPAR expression
100
B
1600
GluR1
A
Subunit
% of total AMPAR expression
100
90
80
70
60
50
40
30
20
10
GluR8
GluR5
GluR4
GluR3
GluR2
GluR1
0
Subunit
Fig. 3.10. Percent composition of AMPAR subtype subunits. The percent composition
of AMPAR was not altered in AII compared to Mature in A) PVC sensory neurons B)
BSC sensory neurons and C) whole abdominal ganglion.
one-way ANOVA on a logistical regression, p>0.05 for all comparisons
Chapter 4:
Transcriptional changes during aging in sensory neurons innervating TWR in
Aplysia
4.1 Summary
Whole-transcriptome expression changes in sensory neurons of aging Aplysia
were studied in two cohorts at maturity and advanced age. Behavioral aging of both the
righting and tail withdrawal reflexes was confirmed in 2 cohorts of animals. Sensory
neurons located in the pleural ventral caudal region of the pleural ganglion that control
tail withdrawal were then tested for differential expression of genes between maturity and
advanced age using the technique known as RNASeq. Many genes and gene ontology
categories that were significantly altered in aged animals compared to maturity code for
genes related to ion channels and neuronal function. This implies that aging has a
profound effect on PVC neurons at the transcriptional level and corroborates
physiological experiments that found reduced PVC neuron function with age. Several
iGluR subunits were significantly downregulated in aged PVC neurons, corroborating
qPCR evidence in Chapter 3. Many upregulated genes in Aged II PVC neurons are
associated with stress response and defense against oxidative damage. Increased stress
response during aging has been characterized in many other species and thus may be a
ubiquitous effect on the aging nervous system of all animals.
4.2 Background
Aging is a biological process resulting in progressive declines in physiological
function, and is the result of a multitude of changes at the cellular and molecular levels.
81
82
Accumulation of reactive oxygen species (ROS) has been proposed to play a
major role in age-related loss of function via oxidative damage. ROS are oxygen-based
radicals produced as a byproduct of cellular metabolism (Braunersreuther and Jaquet
2012; Nayernia, et al. 2014). In healthy neurons, scavenging molecules and enzyme
levels match ROS production, preventing ROS accumulation and oxidative damage to
neurons (Circu and Aw 2010). During aging, increased ROS production and decreased
clearance can lead to oxidative stress that contributes to DNA damage and cellular
apoptosis observed in aging and other neurodegenerative diseases (Uttara, et al. 2009).
Accumulation of ROS in the mitochondria leads to significant damage to the
mitochondrial membrane, further exacerbating ROS production (Shigenaga, et al. 1994;
Toescu, et al. 2000; Floyd and Hensley 2002).
Large-scale gene expression changes occur in the nervous systems of both
vertebrates and invertebrates during aging. In the human cortex, genes that mediate
synaptic plasticity such as protein kinase C, synapsin, and calmodulin had reduced
expression throughout aging (Lu, et al. 2004). These researchers suggested that DNA
damage in promoter regions of these genes may underlie reduced expression. Reduced
expression of genes involved in synaptic plasticity and neuronal function has also been
described in Rhesus macaques, rats, and mice (Bishop, et al. 2010). Increased stress
response in aged animals appears to be ubiquitous and has been characterized in many
species (Yankner, et al. 2008). In the invertebrate model Drosophila melanogaster, both
oxidative stress and aging are characterized by similar changes in gene expression,
including upregulation of stress response genes such as heat shock proteins and
antioxidant genes (Landis, et al. 2004).
83
In addition to transcriptional changes, alterations in neuronal morphology and
physiology have also been implicated in age-related cognitive impairment in the
vertebrate nervous system. Age-related performance decreases in tests of the memory of
the location of a reward in Rhesus macaques may be due to loss of dendritic spines on
glutamatergic pyramidal neurons and decreased density of synapses in the prefrontal
cortex (Dumitriu, et al. 2010). Declines in recognition memory in aged rats has also been
associated with reduced density of dendritic spines (Wallace, et al. 2007). Rhesus
macaque displayed reduced frequency of glutamate receptor-mediated postsynaptic
currents in pyramidal cells that has been linked with impaired memory performance
(Luebke, et al. 2004). Impaired working memory performance in aged rats was
significantly correlated with dopamine deficiencies in the frontal cortex (Luine, et al.
1990).
Cumulatively, morphological, physiological, and transcriptional changes with age
can result in whole organism declines in learning and memory, locomotion, and reflex
systems (Chung, et al. 2005; Kumar and Foster 2007; Bordner, et al. 2011; Kempsell and
Fieber 2015b) Using vertebrate model systems, behavioral deficits with age can be
attributed to neuronal changes in specific brain regions. However, analysis of changes in
the exact neurons underlying behavioral deficits is difficult due to the complexity of the
mammalian brain and complicated neural networks.
Aplysia californica (Aplysia) is a marine mollusk that has been used extensively
in research correlating physiological, behavioral, and molecular changes in neuronal
circuits during aging (Rattan and Peretz 1981; Peretz, et al. 1982; Peretz, et al. 1984;
Kandel 2001; Liu, et al. 2011; Kempsell and Fieber 2014, 2015b). Due to a less complex
84
nervous system than vertebrates, behavioral alterations in aged Aplysia can be studied at
the physiological and molecular levels in the individual neurons that control the
behaviors.
The siphon-gill withdraw reflex (GWR) is one such behavior in which agedrelated changes in behavior have been correlated with physiological and molecular
alterations in the underlying neural circuit. In GWR, a weak stimulus applied to the
siphon causes the siphon and gill to withdraw into the mantle for protection (Pinsker, et
al. 1970). This circuit consists of 24 mechanosensory neurons that make monosynaptic
connects to six motoneurons in the abdominal ganglion (Byrne, Castellucci, Carew, et al.
1978). In aged animals, the reflex shows reduced responsiveness to low intensity stimuli
and altered habituation compared to young animals (Rattan and Peretz 1981). These
behavioral deficits with age have been linked to alterations in morphology of the
neuromuscular junction in the motoneurons of GWR, as well as decreased input
resistance and decreased size of post synaptic potentials evoked by gill stimulation, both
of which reduced excitability and synaptic communication (Peretz, et al. 1984). These
studies on GWR provided evidence that behavioral perturbations with age were a
consequence of alterations in the precise neurons that innervate the reflex.
Recent molecular work on aging of GWR has focused on the L7 motoneuron, one
of the six responsible for moving the tail. L7 is one of the largest neurons in the animal
kingdom and contains sufficient RNA for single neuron molecular studies. High mRNA
abundance of amyloid precursor protein (APP) was observed in L7 using in situ
hybridization (Moroz and Kohn 2010). Increased APP production with age has been
associated with neurodegenerative disease in mammals, including humans (Hardy and
85
Selkoe 2002). In healthy neuronal tissue, production of APP is matched by its clearance
from the tissue (Castellano, et al. 2011). In aging, however, increased expression of APP
in brain tissue and reduced clearance leads to deposition of amyloid beta plaques.
Amyloid plaques reduce synaptic connections and overall neuronal function, causing
memory impairments (Shankar, et al. 2008). Thus, APP in aged Aplysia may lead to
buildup of amyloid plaques in L7, contributing to decreased function and synaptic
transmission between L7 and pinnule muscles of GWR (Peretz, et al. 1982). APP was
also found to be highly expressed in other motoneurons in the abdominal ganglion, as
well as in R2, a cholinergic neuron (Moroz and Kohn 2010). Thus, APP production may
be a consequence of aging in many parts of the Aplysia nervous system that contributes
to reduced neuronal function with age.
Other genes with altered expression in L7 of aged Aplysia include transcription
factors and transcriptional regulators CREB1, CREB2, and S6K (Kadakkuzha, et al.
2013). Transcription factors are proteins that bind to specific DNA sequences and are
responsible for controlling the rate of gene transcription of many other genes. Alterations
in the expression of transcription factors may in turn affect gene expression of other
downstream genes. Hence, changes in transcription factor abundance may be a
contributing factor in age-associated changes in gene expression. Aging in the nervous
system of Aplysia has been shown to result in ~10% of transcripts with ³ 2 fold change
in expression (Kadakkuzha, et al. 2013), which may be due in part to transcription factor
regulation. Together, these studies have shown that age-related alterations in GWR
behavior are correlated with a multitude of changes in morphology, physiology, and
transcription in the underlying neuronal circuitry.
86
Physiological and molecular correlates of aging have also been described in R15,
a bursting neuron in the abdominal ganglion. R15 is believed to be involved in egglaying. Alterations in acetylcholine-induced action potentials occurred in R15 during
aging, as well as reduced expression of some acetylcholine receptors (Akhmedov, et al.
2013). Transcriptional changes in biological pathways for cell signaling, cellular
function, and neurological diseases in aged R15 neurons correlated with observed
physiological declines (Kadakkuzha, et al. 2013). Although there are ample physiological
and molecular data with aging, the behavioral correlates for these changes is an
understudied area of research.
In this study, a different, well-defined neuronal circuit in Aplysia, the tail
withdrawal reflex (TWR), was used. TWR shows predictable behavioral and
physiological aging. Direct tail stimulation initiating TWR activates mechanosensory
neurons in the pleural ventral caudal region of the left and right pleural ganglion (PVC).
PVC neurons make monosynaptic connections to three motoneurons of the pedal
ganglion (P7-P9), which release neurotransmitter at the neuromuscular junction to trigger
movement of the tail muscle (Walters, et al. 1983a). Behavioral studies have shown
characteristic, predictable increases in TWR time to completion as the animal ages
(Kempsell and Fieber 2014). Aplysia can be classified as Mature (M), Aged I (AI), and
Aged II (AII) based on several factors: time since the onset of maturity, diminished reflex
behaviors, as well as physiological changes, discussed below.
Behavioral aging of TWR in AII animals is correlated with physiological changes
in both PVC sensory neurons and pedal motoneurons of the reflex. AII PVC sensory
neurons exhibited decreased frequency and density of currents induced by the both L-
87
glutamate (L-Glu) and D-aspartate (D-Asp) compared to M (Fieber, et al. 2010; Kempsell
and Fieber 2014). AII pedal motoneurons of TWR exhibited reduced excitatory
postsynaptic potential amplitude (EPSP) after tail touch compared to M (Kempsell and
Fieber 2015b). Together, reduced excitability of PVC sensory neurons and decreased
EPSP amplitude in PVC motoneurons suggest that synaptic transmission is negatively
affected by age. Reduced synaptic transmission contributes to aging of TWR and
increased times to complete the reflex in AII animals.
Alterations in sensitization, a form of nonassociative learning, also occurred in
TWR of AII animals. During sensitization, repetition of a stimulus results in a
progressive amplification of the response via serotonin (5-HT) released by facilitory
interneurons onto the sensorimotor synapse (Walters, et al. 1983b; Glanzman, et al. 1989;
Mackey, et al. 1989). 5-HT application to TWR neurons resulted in increased excitability
of PVC sensory neurons and increased transmission between PVC neurons and tail
motoneurons (Kempsell and Fieber 2015b). In M animals, repeated tail shocks also
produced sensitization of the reflex, via increased amplitude and duration of TWR to a
subsequent stimulus (Walters, et al. 1983b). In AII animals, tail shocks and 5-HT
application failed to increase excitatory responses in PVC neurons, and EPSP amplitude
in tail motoneurons was significantly smaller compared to M (Kempsell and Fieber
2015b). Thus, reduced facilitation of TWR in AII is observed the level of both behavior
and the neural circuit.
Studies on the behavioral and physiological correlates of aging of TWR have
greatly enhanced our understanding of the neuronal mechanisms that underlie diminished
responses and reduced synaptic facilitation in AII. Diminished excitability of PVC
88
sensory neurons and reduced EPSP amplitude evoked in tail motoneurons both contribute
to aging of TWR. It has been proposed that decreased EPSP in tail motoneurons may be
the direct result of decreased excitability in PVC sensory neurons (Kempsell and Fieber
2014). The cause of altered PVC function in AII is a critical question that can provide
insights into aging of TWR.
The goal of this study was to characterize molecular correlates of altered
physiology and behavior in aged sensory neurons innervating TWR. Wholetranscriptome changes were analyzed in PVC sensory neurons from sexually mature and
old Aplysia, using RNASeq. First, behavioral experiments were conducted to confirm
aging of TWR. Correlates of aging, including behavioral assessments, were compared
against Kempsell and Fieber (2014) to designate the stage of aging of age 12 month
animals as AII. Then, RNASeq was used to analyze whole-transcriptome changes in
expression in PVC of M and AII Aplysia.
4.3 Materials and Methods
Animal rearing and behavioral assessments of aging
Two cohorts of Aplysia, that originated from differently parented egg masses of
wild-caught animals, were reared at the University of Miami National Resource for
Aplysia. Both cohorts were maintained at a maximum of 5 per cage and fed an ad lib diet
of Agardhiella subulata throughout their life as previously described (Gerdes and Fieber
2006).
Righting and TWR assessments were executed beginning at M on 6-8 randomly
selected animals from each cohort using previously described protocols (Chapter 3). M
89
was defined as the day the first egg mass was found and occurred at age 8 months in both
cohorts. To quantify behavioral aging, behavioral assessments were continued monthly
on randomly selected animals (n=6-8) from each cohort through the end of life, age ~12.5
months in both cohorts.
RNA extraction and sequencing preparation
RNA was extracted from PVC neurons from six animals from each cohort at M
and six animals from each cohort at AII. The animals chosen for dissection of ganglia and
RNA extraction from neural clusters were a randomly selected subset of the animals
tested for TWR and righting at M and AII. This allowed for direct correlation of an
individual animals’ behavioral assessments with its expression profile.
To remove PVC neurons for RNA extraction, animals were first anesthetized by
injection of 1/6th body weight of isotonic MgCl2. The CNS was then removed by severing
the connectives and removal of all nervous system ganglia. PVC neurons were separated
from the remainder of the pleural ganglion by pinning the pedal and pleural ganglia
tightly in a sylguarded 35mm plastic dish filled with artificial sea water (ASW: 417mM
NaCl, 10mM KCl, 10mM CaCl2, 55mM MgCl2, and 15mM HEPES-NaOH, pH 7.6).
PVC neurons were peeled away from the remainder of the pleural ganglion with a pair of
fine forceps after removal of the surrounding connective tissue. PVC neurons from both
hemiganglion were pooled as a single sample.
Total RNA was extracted with the Qiagen RNeasy Micro Kit (Cat. #74004)
following the manufacturer’s instructions. Samples were treated with DNase to remove
any contaminating DNA. Purified RNA samples were stored in 2.5 volumes of ethanol
with 0.3M sodium acetate at -20°C until further processing. Samples in ethanol and
90
sodium acetate were spun in a microfuge at 4°C at full speed for 20 minutes, ethanol
removed, and the RNA pellet resuspended in 15µl of TE buffer before library
preparation. RNA quantity and quality were assessed with both a Nanodrop (model ND1000) and an Agilent 2100 Bioanalyzer prior to library preparation.
**Two hundred ng of RNA from the 3 highest quality RNA samples from each
cohort at M and AII were used in downstream library preparation. The highest quality
RNA samples were defined as having sufficient RNA quantity and 260/280nm ratio ~2 to
ensure no DNA or protein contamination. Libraries were prepared using the Illumina
TruSeq Stranded Total RNA Low-Throughput Library Prep Kit (Cat. #RS-122-2201)
following to the manufacturer’s instructions. Beckman Coulter Agencourt RNAClean XP
(cat# A63987) and Beckman Coulter Agencourt AMPure XP (cat# A63881) beads were
used for recovery of RNA and DNA, respectively. Following library preparation, samples
were assessed on the bioanalyzer for quantification and to verify 200-300bp fragment
sizes. Libraries were sequenced as 100bp paired-end reads by Elim Biopharmaceuticals
(Hayward, CA) in one lane on a HiSeq 2500 high-throughput sequencer (Illumina).
Annotation information
Aplysia exon information (ref_AplCal3.0_gnomon_scaffolds.gff3) and genome
annotation (version AplCal3.0 from July, 2015) were downloaded from the NCBI ftp
server database.
Data processing
Raw reads were quality filtered using the fastaxtoolkit
(<http://hannonlab.cshl.edu/fastx_toolkit/>). When a base pair quality score of <20 was
encountered in a read, the remainder of the read was cut off and removed. Reads shorter
91
than 20 base pairs after trimming and removal of barcodes were also removed. Quality
score and nucleotide distributions for each library were visualized using the Galaxy web
server (Blankenberg, et al. 2010).
rRNA was removed during sequencing preparation, however M sample #6 still
contained many rRNA reads in its library. In order to assess an accurate number of total
reads used for downstream analysis it was necessary to remove rRNA reads from this
library prior to analysis. rRNA reads were removed by mapping all reads to the Aplysia
rRNA annotation with the STAR aligner (parameters described below) and discarding
reads that aligned to the rRNA annotation. Therefore, downstream analysis did not
contain rRNA reads.
Mapping Reads to the genome with Tophat2 and STAR
After quality filtering, reads were mapped to the Aplysia genome using both the
Tophat2 and STAR aligners. The indexed genome for use with Tophat2 was built using
Bowtie 2.2.6.0 (Langmead and Salzberg 2012) with default parameters from the
AplCal3.0 genome fasta file. The quality controlled reads were then aligned using
Tophat2 v2.1.0 (Kim, et al. 2013) to the indexed genome. Additional options beyond the
default that were used during alignments were --no-discordant and changes to --mateinner-dist, as described below.
The –no-discordant option removes discordant pairs from the alignments. While
some of the discordant pairs may be true fusion transcripts, the Tophat2 authors found
that most were mis-mapping artifacts. Because fragment lengths used for sequencing
ranged from 250bp to 300bp, the alignments were run with the expected distance between
mate pairs at both 50bp and 100bp. We found that the number of uniquely mapped reads
92
increased ~1-2% with the mate pair distance at 50bp in 3 libraries, and this option was
then used for the remaining libraries. Uniquely mapped reads were assigned to Aplysia
RefSeq transcripts using featureCounts (Liao, et al. 2014).
Genome index files for use with STAR were generated using the --runMode
genomeGenerate option using both the Aplysia AplCal3.0 genome and the
ref_AplCal3.0_gnomon_scaffolds.gff3 transcript annotation. Quality controlled reads
were aligned using the STAR aligner (Dobin, et al. 2013). The --quantMode Genecounts
option was utilized to count the number of reads uniquely mapping to each transcript
using the HTSeq-count program (Anders, et al. 2014). Mapping and counting for pairedend and singletons were executed separately, and total counts for each transcript was
determined by summing the counts for the paired and singletons reads for each sample.
Differential expression analysis
Statistical testing for differential expression (DE) of transcript counts was
performed using DeSeq2 (Love, et al. 2014), a method based on the negative binomial
distribution and performed in the R statistical environment (Team 2014). Only reads that
uniquely mapped to the genome were counted and used for analysis. Reads that were
expressed in at least two biological replicates were analyzed for DE. Raw read counts
were normalized in DeSeq2 to adjust for differences in library sizes. Significant DE was
defined as p£0.05 after false discovery rate (FDR) correction. Data were normalized
using the regularized log transformation for principal component analysis (PCA).
Of the top 50 most significantly DE genes, 17 were classified as uncharacterized
proteins. Each uncharacterized protein was analyzed in a BLAST search against human
proteins and for conversed domains in InterPro (Hunter, et al. 2011) to attempt to
93
annotate these genes. Three genes were annotated in this way: CCAP1, DNA ligase-like,
and glycine N-acyltransferase-like. None of these three genes were previously annotated
in the Aplysia transcriptome database.
Gene ontology enrichment
Transcripts found to be DE were first translated into their corresponding protein
sequences prior to gene ontology analysis. Translated sequences were blasted against the
Homo sapiens refseq protein database (version 9606.9558) using Blast2GO (ver. 3.3, Evalue£1.0E-3), to identify proteins with human homologs for gene ontology (GO)
analysis (Götz, et al. 2008). This allowed for mapping of DE genes to the better-studied
and annotated human GO pathways. GO annotation terms with E-value≤1.0E-6 were
used in downstream analysis. This process was repeated for the full Aplysia
transcriptome (AplCal3.0) as a reference set for enrichment analysis.
GO terms of proteins translations from DE transcripts were tested against the
reference set of GO terms for the full Aplysia genome annotation to test for biological
processes and molecular functions that were over-represented (enriched). Using a
Fisher’s exact test, GO terms that were over-represented at an FDR corrected pvalue£0.05 were identified.
We were particularly interested in the effects of aging on neuroexcitability, and
we found many GO categories involved in ion channel activity. To reduce redundancy in
these very similar categories, these were combined into a single GO category during the
analysis which we called “ion channel associated proteins.” P-value for ion channel
associated proteins was calculated by using all unique genes in both the DE genes and the
full genome with ion channel function and a Fisher’s exact test in R. To calculate an FDR
94
corrected p-value for this category, p-values from all GO categories from DESeq2
analysis were FDR corrected using the fdrtools package in R (Strimmer 2008).
Categories combined for ion channel associated proteins are highlighted in the complete
list of enriched GO terms (Supplementary File 1).
qPCR validation of selected genes
qPCR was used to verify the expression status of selected genes identified from
the RNASeq analysis. For these experiments, different animals from the same two
cohorts as the RNASeq experiment were used. Four animals at M and four animals at
AII, thus 8 total animals, were tested for the righting and TWR, then sacrificed for RNA
extraction as previously described for RNASeq. RNA was quantified on a Nanodrop
(model ND-1000), and 100 ng of RNA was reverse transcribed to cDNA using the
SuperScript III First-Strand Synthesis System with random hexamer primers (Invitrogen).
The resulting cDNA was diluted 1:5 with nuclease-free H2O to provide the working
concentration of cDNA.
Primers designed for each gene were 18-22 bp and generated target amplicons of
75-125 bp. Primer sequences and amplicon length for each gene can be found in Table
4.1. To assess amplification of the desired target sequences, each set of primers was
tested by PCR amplification of cDNA. PCR products were separated on a 1% agarose gel
to confirm the appropriate size for each amplicon. These bands from the agarose gel were
then cut out and the DNA purified using a Qiagen Gel Purification kit and purified gel
products were sequenced by Genewiz (South Plainfield, NJ) to confirm amplification of
the desired target sequence.
95
qPCR reactions were carried out on a Stratagene Mx3005P with SYBR Green
master mix under the following conditions: 95°C for 10 minutes, followed by 40 cycles
of 95°C for 15 sec, 58°C for 30 sec, 72°C for 30 sec. For each biological replicate two
technical replicates were performed. All run were normalized to GAPDH control and
relative standard curves were used for analysis. Expression ratios were calculated for
genes between M and AII, as described in Chapter 3.
4.4 Results
Behavioral changes in TWR and righting with age
There were no significant differences observed in either TWR or righting between
the two cohorts at any age (Student’s t-test, p£ 0.05). Therefore, times were combined
from the two cohorts for analysis of behavioral aging. TWR significantly increased with
age in both cohorts (Fig. 4.1A; 14.0±0.4 sec at age 8-10 mos (mean±SE), 19.6±0.6 sec at
11 mos, and 26.4±0.7 sec at 12 mos; one-way ANOVA p£ 0.05; Tukey’s post-hoc
analysis). Time to right also increased significantly with age (Fig. 4.1B; 13.6±0.4 sec at
age 8-10 mos, 17.1±0.5 sec at age 11 mos, 20.1s±0.5 sec at age 12 mos; one-way
ANOVA p£0.05; Tukey’s post-hoc analysis). Age>11 months, TWR time>20.1 sec, and
righting time > 18 sec in age 12 month animals of this study corresponded to AII based
on previously defined stages of aging in Aplysia (Kempsell and Fieber 2014).
Read depth and mapping
After quality control filtering, a total of ~210 million reads were used for
downstream analysis (an average of ~17.5 million reads/individual). Quality score
96
boxplots of the reads showed that base quality scores were high for all 100bp, indicating
high quality sample preparation and sequencing (data not shown).
Genomic mapping of reads was performed using both STAR and Tophat2. 5070% of reads per individual uniquely mapped to the NCBI Aplysia californica
Annotation Release 101 genome using either aligner (assembly AplCal3.0). In all
samples STAR yielded a higher percentage of uniquely mapped reads compared to
Tophat2 (~5-10%, data not shown). A summary of read depth and mapping for each
library with the STAR aligner can be found in Table 4.2.
Identifying DE genes in PVC neurons with age
RNA from PVC neurons from six biological replicates at M and AII were isolated
and analyzed for DE. The correlation between the biological replicates within each group
was calculated with a principal component analysis (PCA) to determine if overall gene
expression patterns differ between M and AII animals (Fig. 4.2). This analysis showed
that the first two principal components accounted for 57% of the variability in the dataset.
Most biological replicates clustered together, indicating that expression profiles were
most similar between individual M animals on the one hand and individual AII animals
on the other. One exception was a single M animal that showed an intermediate
expression profile that did not clearly belong to either M or AII (Fig. 4.2, circled).
There were 1202 and 1132 DE genes identified with STAR and Tophat2,
respectively. Of these DE transcripts, >80% were identified with both analyses (Fig. 4.3).
Due to high agreement between the methods, downstream gene ontology analysis used
only the STAR aligner data. A full list of DE transcripts from STAR at p≤0.05 can be
found in Supplementary File 2.
97
In aging, approximately as many genes were upregulated as were downregulated
in PVC neurons (Fig. 4.4). Six hundred sixty-five of the 1202 DE genes (55.3%)
increased expression in AII animals, with 537 genes (46.7%) having decreased
expression. A heatmap of the top 50 genes with the lowest p-values can be found in Fig.
4.5. Several of the genes in this list that exhibited decreased expression in AII animals
play a role in neuronal excitability, including voltage-gated potassium channels and
ionotropic glutamate receptors (iGluR). Other genes with decreased expression are
involved in induction of long-term memory, including adenylate cyclase (Fig. 4.5) and
the catalytic subunit of protein kinase A (PKA; Supplementary File 2). Several transcripts
that were upregulated in AII play a role in cell protection and oxidative stress response,
including major vault protein, heat shock protein, and multi-drug resistance protein.
To determine if batch effects from using two cohorts played a role in DE
expression, DE was analyzed between the two cohorts, without accounting for age. We
found 25 DE transcripts between the two cohorts (p≤0.05, FDR corrected). None of these
25 genes were also DE in the aging analysis. This is a very low number of transcripts
compared to >1200 DE transcripts between age groups, and indicated that there were few
differences in gene expression due to the biological variability of the two cohorts.
Therefore, differences in expression in this study are likely attributed to transcriptional
changes between M and AII.
Expression of iGluR subunits
iGluR subunits were analyzed to test if reduced expression of iGluR coincided
with previously observed declines in L-Glu mediated responses of AII PVC neurons.
STAR was unable to differentiate between the two splice variants of Grin1, Grin1-1 and
98
Grin1-2, therefore Grin1 expression in this study encompasses both splice variants. There
were no changes in expression of NMDAR subunits Grin1 or Grin2 (Fig 4.6). GluR2,
GluR3, belonging to the AMPA subtype, and KA2, a kainate subunit, were significantly
down-regulated in AII PVC neurons compared to M PVC neurons. GluR4 and KA1
subunits were not analyzed due to a low number of counts for these two genes. These
results are similar to qPCR results of Chapter 3, in that some iGluR genes were downregulated in AII PVC neurons, and zero subunits were upregulated (See Figs. 3.4 and 3.5,
in Chapter 3). However, all down-regulated iGluR genes observed with RNASeq were
different than those found using qPCR.
The percent contribution of each subunit to the total expression for NMDAR and
AMPAR subtypes was calculated to determine if receptor makeup may have changed
during aging. There were no statistically significant changes in the percent contribution
of either Grin1 or Grin2 NMDAR subunits between M and AII (Fig. 4.7A). In AMPAR
subunits, GluR1 comprised a significantly greater percentage of the total AMPAR
expression in AII than at M, and GluR3 contributed a significantly smaller percentage to
total AMPAR expression (Fig. 4.7B). These findings largely corroborated the aging
results of Chapter 3, where no significant changes in subunit composition were found.
The apparent substitution in AMPAR of GluR3 for GluR1 found here was the exception.
Gene ontology analysis
DE genes were mapped to gene ontology pathways using Blast2GO to determine
the biological processes and molecular functions that were most significantly affected
during aging. BLAST hits matching human proteins were identified for 1054 of the 1202
DE Aplysia proteins (E-value<1E-3). Approximately 1000 gene ontology terms were
99
found to be over-represented in the analysis (p£0.05, FDR corrected), suggesting that
many functions were significantly altered in AII animals. A complete list of enriched
ontology terms can be found in Supplementary File 1.
Fig. 4.8 presents the most significantly enriched GO categories. The GO category
ion channel associated genes was the most significantly enriched molecular function in
PVC neurons (p£0.05, Fig. 4.8). Furthermore, >80% of DE ion channel associated genes
showed reduced expression in aged animals, suggesting decreased ion channel function
occurred in aged animals (Fig 4.9). Synaptic transmission was also a significantly
enriched biological process, further emphasizing alterations in neuronal transmission in
AII (Supplementary File 1).
Response to stress was also found to be an enriched biological process
(Supplementary File 1). Approximately 70% of stress response genes DE showed
increased expression in aged animals, consistent with findings of increased stress
response with age in many other species (Fig. 4.10, Bishop, et al. 2010).
qPCR verification of RNASeq
A subset of nine genes important to neuronal excitability in humans were
quantified via qPCR at M and AII, using new RNA samples from the same two cohorts
sampled for RNASeq. Six of the 9 genes selected for qPCR showed significant DE in
RNASeq (Fig. 4.11). Six of the genes coded for ion channel proteins: voltage-gated
potassium channel Shaw (Shaw), voltage-gated potassium channel Shaker (Shaker),
voltage-dependent calcium channel subunit alpha-2/delta-3 (Ca2+ subunit alpha), and
sodium channel alpha-subunit SCAP1. CREB3 was chosen for its role in cell
100
communication and calcium ion transport. CCAAT/enhancer-binding protein (C/EBP) is
a transcription factor important for transcriptional regulation of many genes.
Three other genes tested via qPCR, CREB2, amyloid protein b, and huntingtin
were not found to be DE in RNASeq, but were included for qPCR as genes previously
tested in Aplysia nervous tissue. Two genes had significantly different expression in both
RNASeq and qPCR: Shaw and Ca+ subunit alpha (Fig. 4.11, p£0.05). There were no
significant changes in expression for the remaining 7 genes via qPCR.
4.5 Discussion
Aplysia is an emerging model of aging in the nervous system due to a relatively
compact life span of one year and well-mapped neural circuits (Frazier, et al. 1967;
Carew, et al. 1981; Rattan and Peretz 1981; Stommes, et al. 2005). The use of two
cohorts of hatchery reared Aplysia provided the advantage of studying 2 families of
sibling animals with reduced biological variability compared to wild caught animals.
PVC neurons are the primary source of sensory information for TWR, and their relative
homogeneity makes them ideal for studies of aging in the nervous system.
Many Aplysia genes have homologs in higher vertebrates, making it a relevant
model for age-related neuronal changes that may be important in aging vertebrate brains
(Moroz, et al. 2006). Furthermore, Aplysia contain homologs for many genes that are
relevant to neurological diseases associated with age in humans (Moroz and Kohn 2010).
We found that >85% of the DE transcripts aligned to human transcripts in BLAST
searches (E-value£1E-3). This allowed for mapping most DE genes to the much bettercharacterized human gene ontologies for pathway enrichment, and allowed for
101
predictions of pathways that may be important in aging vertebrates. Another similarity
we found with the vertebrate brain that became an advantage was the number of
expressed transcripts. In mice, ~11,000-13,000 genes are expressed per brain region
(Ramsköld, et al. 2009). We found that a similar ~10,000 transcripts in PVC neurons
expressed at least 5 copy per biological replicate.
Previous transcriptome studies on aging in Aplysia and other model organisms
have found genes to be both up- and downregulated with age (Lee, et al. 2000; Moroz
and Kohn 2010; Kadakkuzha, et al. 2013). Thus, it was unsurprising that aging of PVC
neurons exhibited bidirectional changes in gene expression, with ~55% of DE genes
displaying increased expression in Aged II. This is approximately the same percentage of
genes up-regulated in the R15 bursting neuron of the Aplysia abdominal ganglion
(Kadakkuzha, et al. 2013).
In AII Aplysia, reduced performance of aged PVC sensory neurons has been
proposed to significantly contribute to reduced TWR behaviors. PVC neurons exhibited
reduced excitability to both tail taps and intracellular current injection in AII animals
compared to M animals (Kempsell and Fieber 2014). We examined GO categories to
functionally profile genes that were DE in AII PVC sensory neurons, and found that
reduced excitability coincided with DE of excitation-related genes and gene ontology
categories that underlie the behavior, as follows:
The most significantly enriched molecular function GO category was ion channel
associated proteins, with 87 out of 619 genes in this category exhibiting altered
expression in AII PVC neurons. This result suggested that the aging process in sensory
neurons has profound impacts on regulation of neuronal physiology. Furthermore, despite
102
overall bidirectionality of all gene expression changes in AII PVC neurons, 83% of DE
ion channel associated genes were down-regulated, indicating that ion channel function is
reduced in AII. Ten of the ion channel associated genes encode for voltage-gated
potassium channel proteins, all of which were downregulated in AII PVC neurons.
Reduced expression of voltage-gated K channels Shaw and Shaker were also examined
using qPCR. Shaw showed significantly reduced expression in qPCR, while Shaker did
not.
During an action potential, voltage-gated Na+- and Ca2+ channels in the plasma
membrane open to depolarize excitable cells such as neurons. Depolarization opens a
broad assortment of voltage-gated K+ channels with a delay, and causes outward
movement of K+ ions until the neuron is restored to its resting potential. Ca2+ and K+
channel activation are thus key to the reference level excitability of each neuron. While
ion channel abundances were not measured directly here, mRNA abundance is often
employed as a proxy for amount of protein, and thus receptor number (Vogel and
Marcotte 2012). Therefore, our results that voltage-gated K+ and Ca2+ channel mRNAs
were decreased may mean fewer of these channels in the membrane. A reduction in both
could produce the reduced excitability demonstrated in AII PVC neurons (Kempsell and
Fieber 2015b). Fieber et al (2010) did not observe reduced Ca2+ current density in aged
PVC neurons studied with whole cell voltage clamp. Thus, reduced voltage-gated Ca2+
channel expression found here may only be a minor contributor to reduced excitability
that was undetectable by whole cell voltage clamp. This may suggest that reduced K+
channel function is the major contributor to reduced excitability in AII PVC neurons.
103
Reduced excitability during aging has also been described in several other
neurons of the Aplysia nervous system. In sensory neurons of the buccal ganglion, DAsp- and L-Glu-induced currents declined significantly with age (Fieber, et al. 2010;
Kempsell and Fieber 2014). In L7 motoneurons, spiking was depressed and postsynaptic
potential amplitude was decreased in response to siphon stimulation in aged animals
(Rattan and Peretz 1981). The R15 cholinergic neuron in the abdominal ganglion showed
reduced responsiveness to acetylcholine in aged animals (Akhmedov, et al. 2013).
Together, these results suggest that reduced excitatory responses are prevalent in the
aging Aplysia nervous system in both sensory and motoneurons. Reduced neuronal
excitability also occurs in the aging mammalian hippocampus and is correlated with
memory declines (Barnes 1990; Barnes, et al. 1992; Foster and Norris 1997; Jouvenceau,
et al. 1998; Kumar and Foster 2007). Declines in neuronal excitability appear to be
widespread, and are a hallmark of aging in the nervous system of both mammals and
Aplysia.
In addition to reduced excitability to tail tap and intracellular current injection,
AII PVC neurons also exhibited reduced density of L-Glu-induced excitatory currents
(Kempsell and Fieber 2014). Three iGluR subunits, GluR2, GluR3, belonging to the
AMPA subtype, and KA2, a kainate subunit, showed significantly reduced expression in
AII PVC neurons compared to M via RNASeq. There were zero iGluR subunits with
significantly increased expression in AII. This suggested that reduced L-Glu current
amplitude in AII PVC sensory neurons (Kempsell and Fieber 2014) may be due to
decreased expression of iGluR. These findings further support results in Chapter 3 using
qPCR, where several iGluR subunits had significantly reduced expression in AII PVC
104
neurons compared to M, and zero subunits were significantly upregulated (see Fig 3.4, in
Chapter 3). The iGluR subunits observed to be DE in this study are not the same subunits
that were DE in Chapter 3. This could be due to using different cohorts of animals in this
study, or evidence that transcriptional changes in iGluR vary during aging. Regardless,
both methods confirm reduced expression of iGluR occurs in AII PVC and may underlie
reduced L-Glu-induced excitability in AII.
In Chapter 3, Grin1-1 was shown to be expressed at significantly greater copies
than any other iGluR subunit in PVC sensory neurons, present at a minimum of 2-fold
higher copy number than any other iGluR subunits. The STAR aligner was unable to
differentiate between the two splice variants of Grin1 in the study described here, thus
Grin1 expression encompasses both Grin1-1 and Grin1-2. Grin1 subunits were expressed
approximately three times higher than any other subunit. These findings are confirmatory
of the expression status of Grin1 found in qPCR as the largest contributor to iGluR
mRNA expression in PVC neurons. Further, these independent confirmations of Grin1 as
overwhelmingly dominant strongly suggest that L-Glu-induced excitatory currents in
PVC neurons are mediated by Grin1. Grin1, however, was not DE in this study, although
Grin1-1 and Grin 1-2 were significantly DE in aging in the qPCR results of Chapter 3. It
is possible that maintenance of Grin1 is necessary for neuronal function, and aged PVC
neurons are attempting to compensate for loss of iGluR function by maintaining Grin1
expression levels. We did not test the L-Glu physiology of AII animals in qPCR and
RNASeq cohorts, and instead used behavioral proxies. Declines in physiology with age
occur at different rates in different cohorts of Aplysia (Fieber, et al. 2010; Kempsell and
Fieber 2014). The cohorts studied in chapter 3 with qPCR may have been at a more
105
advanced stage of aging that was unable to compensate with increased Grin1 expression.
Detecting this later stage of aging may not have been possible with behavioral proxies.
A consequence of aging is increased oxidative damage from ROS. Accumulation
of ROS can be detrimental to cell survival by inducing DNA mutations and can disrupt
many cellular processes by altering the structure and function of many macromolecules
(Stadtman 2001). Downregulation of neuronal genes in humans is associated with DNA
damage that can repress transcription (Loerch, et al. 2008). Additionally, oxidation of ion
channels is wide-spread during aging, and can modulate functional changes in voltagegated Na+, K+, and Ca2+ channels (Annunziato, et al. 2002; Patel and Sesti 2016). In
response to oxidative stress, the transcription factor Nrf2 is translocated from the
cytoplasm to the nucleus to bind DNA promoters and increase transcription of stress
response genes (Nguyen, et al. 2009).
Consistent with the idea of increased stress response during aging, GO for
regulation of response to stress (GO:0080134) was significantly enriched in AII PVC
neurons (Supplementary File 1). Increased expression of stress response genes during
aging has previously been described in humans, monkey, rats, mice, fruit flies, and
roundworms (Bishop, et al. 2010). In Aplysia, stress response genes are also upregulated
in several neurons in the abdominal ganglion (Moroz and Kohn 2010; Kadakkuzha, et al.
2013), and therefore stress response is an evolutionarily conserved feature of aging in
many parts of the Aplysia CNS.
Heat shock proteins were upregulated in response to stress in vertebrates, and
work to refold or degrade severely damaged proteins caused by ROS (Murshid, et al.
2013; Rodgers, et al. 2013; Leak 2014). Heat shock protein 70, in particular, protected
106
cells from cell death caused by increased cellular damage in mice (Calderwood, et al.
2009). Four heat shock proteins were upregulated in AII PVC neurons, including heat
shock protein 70, and zero heat shock proteins had decreased expression. Other genes
with increased expression in AII PVC neurons that have been implicated in
chemoprotection and cell survival include major vault protein and multi-drug resistance
protein. Major vault protein has been suggested to play a role in cell signaling and
prevention of stress-induced apoptosis. Increased expression of major vault protein also
occurred in aged humans (Ryu, et al. 2008), and its protein sequence is highly conserved
between Aplysia and humans (Moroz and Kohn 2010). Multi-drug resistance protein has
an important role in response to stressors such as ROS (Eldakak, et al. 2010), and was
also upregulated in aged PVC neurons. Taken together, increased expression of heat
shock protein and other chemoprotective genes in AII PVC neurons may be a
compensatory mechanism in an effort to maintain proper homeostasis by preventing ROS
and other oxidative damage accumulated during aging.
Aplysia has homologs for many genes implicated in neurological diseases and
reduced neuronal function in aged humans (Moroz, et al. 2006). Two of these genes,
amyloid b protein and huntingtin, were tested with qPCR for DE in aged PVC neurons.
Amyloid b is formed by proteolysis of amyloid precursor protein and is the major
component of amyloid plaques. Increased deposition of amyloid plaques in aged neurons
is associated with neuronal impairments related to Alzheimer’s disease (Schenk, et al.
1999). We did not find amyloid b, a derivative of amyloid precursor, to be DE in PVC
neurons with either qPCR or RNASeq. We did, however, find it to be one of the most
highly expressed transcripts in our study. In the abdominal ganglion of Aplysia, amyloid
107
precursor protein was highly expressed in many motoneurons, but DE in only some of
these neurons (Moroz and Kohn 2010). Thus, high expression of amyloid proteins
appears to be prevalent in the Aplysia CNS, but DE is apparently neuron specific, and
was not observed in AII PVC neurons.
Huntingtin is a protein critical for induction of learning in Aplysia (Choi, et al.
2014), and expression of mutant huntingtin induced neurological deficits in Huntington’s
disease mouse models (Nguyen, et al. 2005; Bradford, et al. 2009). We found no change
in huntingtin expression with either qPCR or RNASeq, indicating that altered huntingtin
expression may not be involved in age-related neuronal deficits of PVC neurons.
Together, these results suggest that aging and impaired physiology of PVC neurons in
Aplysia may be due to normal aging and aging in this model may not predict roles for
genes involved in neurological diseases.
Induction of both short and long-term memory is compromised in aged Aplysia,
including in TWR (Bailey, et al. 1983; Kempsell and Fieber 2015b). Adenylate cyclase
and the catalytic subunit of PKA, both necessary for the induction of sensitization in
sensory neurons, were downregulated in AII PVC neurons. During facilitation,
sensitization induces release of 5-HT, which binds to G-protein coupled receptors on the
sensory neurons to activate intracellular adenylate cyclase (Kandel 2001). Activated
adenylate cyclase converts ATP to cyclic AMP (cAMP), and thus increased adenylate
cyclase enhances intracellular cAMP concentrations. cAMP removes the two regulatory
subunits from cAMP-dependent protein kinase A (PKA) to release the catalytic subunit
of PKA. The PKA catalytic subunit closes K+ channels via phosphorylation of the
channel. This, in turn, increases action potential duration, thereby increasing the amount
108
of free Ca2+intracellularly. Increased intracellular Ca2+ results in an increase in
neurotransmitter release at the sensory to motoneuron synapse and increased synaptic
strength characteristic of memory formation.
Direct injection of PKA into AII PVC neurons induced facilitation, suggesting
that processes upstream of PKA activation may be compromised during aging (Kempsell
and Fieber 2015a). Reduced transcription of adenylate cyclase found in this study may
result in less adenylate cyclase proteins available to be activated by 5-HT in aged
neurons. This would in turn have downstream effects on cAMP activation. In addition to
potentially reduced cAMP, reduced availability of the catalytic subunit of PKA could
result in less PKA available for activation by cAMP. With fewer active PKA subunits
phosphorylation of K+ channels would be decreased. Thus, increased action potential
duration and increased excitability necessary for Ca2+ influx and enhanced
neurotransmitter release may be affected. These molecular mechanisms may in part
explain reduced excitability of PVC and reduced EPSP amplitude via reduced activation
of proteins necessary for enhanced neurotransmitter release.
Cognitive declines in memory tasks with age have been described in many species
(Wallace, et al. 2007; Dumitriu, et al. 2010). Decreased expression during aging of genes
involved in synaptic plasticity have been proposed to underlie many of these declines in
mammals (Lu, et al. 2004). Reduced expression of genes involved in learning and
memory systems appears to be a ubiquitous effect of aging in the nervous system of
animals via reduced transcription of components necessary for induction of facilitation.
This study examined whole-transcriptome changes that occur during aging in
PVC sensory neurons. Gene expression changes during aging in Aplysia are variable
109
depending on the neurons studied, and this is the first study to analyze sensory neuron
aging. The results suggested that the aging process in PVC sensory neurons coincides
with decreased expression of many plasma membrane proteins important to excitability,
such as voltage-gated K+ channels, voltage-gated Ca2+ channels, and iGluR. The
summation of altered receptor expression with age may result in reduced PVC neuronal
performance observed in AII Aplysia. Although changes in expression vary in the
Aplysia nervous system and in different species, upregulation of stress response genes
and decreased expression of genes important for synaptic plasticity are evolutionarily
conserved changes with age observed in many vertebrates. Aplysia genes have many
homologs in vertebrates, and conserved molecular pathways of aging and simple neural
circuits together make Aplysia a powerful model for the effects of aging in the nervous
system.
110
Table 4.1. Primer sequences for qPCR.
Gene
CCAAT/enhancer-binding
protein zeta-like
Primer Sequences for qPCR (5’-3’)
F ATCAACTCAGCCGACGTCTC
R GGGACTGCTTCTTCGTCTTGA
Amplicon length (bp)
114
sodium channel alpha-subunit
SCAP1
F ATCGGCATGCAGCTCTACTC
R ACCCAATCGTTCCACTCGTC
104
potassium voltage-gated
channel protein Shaw-like
CREB2
F
R
F
R
F
R
F
R
F
R
F
R
F
R
104
Huntingtin
voltage-dependent calcium
channel subunit alpha-2/delta-3
CREB3B
amyloid beta A4
potassium voltage-gated
channel protein Shaker-like
TGGACATAAAGGCGTGTTGC
GACTGTCATCTGTTCCCGCT
TTGTTGGGTGGCATGGAACT
AGTTCTTCAGCACCGCCAAT
TGGACACTCAGACCACCAGT
CTCTAATAACGCTGCACGGA
ACTTATGAGGGCCTGGGTCT
ACGTCCGTTCCTACAACACC
TGCCAACGAACTTACCCCTC
CCTTTTGCGGCTCTCTTTCG
ATCAGAATAGACGACGCGGG
CTGCCTCTCTCGTAGTCGTG
GCCACTGGGGAAGCTCATAG
CCTCTTCCAGTCTTGTGCGA
107
76
122
101
101
120
111
Table 4.2. Read depth and mapping statistics using STAR. Unpaired reads had a read
removed during quality control.
Sample
Mature1
Mature2
Mature3
Mature4
Mature5
Mature6
Aged1
Aged2
Aged3
Aged4
Aged5
Aged6
paired
Total Number of
reads after
Trimming
19,608,062
% reads mapped
to genome by
STAR
66.76%
unpaired
1,037,824
79.24%
paired
unpaired
paired
unpaired
paired
unpaired
paired
unpaired
paired
unpaired
paired
unpaired
paired
unpaired
paired
unpaired
paired
unpaired
paired
unpaired
paired
unpaired
14,783,156
956,055
13,576256
918,426
14,511,446
627,032
25,116,528
1,502,596
28,406,435
1,886,646
18,443,688
1,113,916
18,051,002
960,838
17,130,158
1,157,305
15,336,780
710,853
14,527,360
656,870
10,089,056
832,514
70.54%
84.06%
72.74%
81.13%
61.82%
60.07%
72.62%
75.38%
50.64%
77.29%
59.36%
77.29%
63.18%
57.74%
60.77%
75.96%
58.36%
45.04%
60.41%
56.38%
52.91%
57.39%
Total # transcripts
counted by htSEQ
3,090,792
2,482,607
2,582,177
2,187,782
4,549,233
3,108,397
2,664,852
2,525,682
2,768,254
1,911,932
1,464,547
1,525,506
112
A
Time to relax tail (s)
30
*
25
*
20
15
10
5
0
8
B
9
10
11
Age (months)
12
25
Time to right (s)
*
20
*
15
10
5
0
8
9
10
11
Age (months)
12
Fig. 4.1. Behavioral assessments of aging in cohorts used for RNASeq. Two cohorts
of Aplysia were tested monthly for behavioral declines associated with aging. For each
behavior there were no significant differences between cohorts during any month,
therefore data were combined from both cohorts at each time point. (A) TWR increased
significantly with age (all data presented as mean ± standard error). B) Righting also
increased significantly with age. Reflex completion times for age 12 month animals used
for RNASeq corresponded to AII based on previously published stages of aging in
Aplysia (Kempsell and Fieber 2014).
*denotes significant difference compared to other ages, p£ 0.05, one-way ANOVA
113
Mature
Mature
Aged
Aged
II
PC2
PC2
1010
00
−10
-10
-20
−20
-10
−10
0PC1
0
10
10
20
20
30
30
PC1
Fig. 4.2. Principal component analysis (PCA) of the first two principal components
in M and AII PVC neurons. Gene expression profiles of M and AII PVC neurons were
plotted against the first two principal components. Clustering of M and AII PVC neurons
separately indicates expression profiles were most similar within each age group. One M
animal exhibited an expression profile that did not clearly belong to either age group
(circled). Colored circles represent 95% confidence interval.
114
Fig. 4.3. Overlap of DE genes identified with STAR and Tophat2. Weighted Venn
diagram comparing the number of genes found to be DE in DeSeq2 (p£0.05, FDR
corrected) after genome alignment with STAR and Tophat2. High overlap indicates good
agreement between the transcripts found to be DE with two aligners.
115
3
log 2 fold change
2
1
0
-1
-2
-3
1e-01
1e+01
1e+03
1e+05
Mean expression
Fig. 4.4. Scatterplot of mean expression of genes analyzed by STAR vs. their log-fold
change. Points in red indicate significant DE at p£0.05. Negative log-fold changes along
the y-axis corresponds to transcripts that showed lower expression in AII animals. There
were approximately equal numbers of genes up and down regulated in AII PVC neurons.
116
Expression
Aged II
Expression
Aged II
Expression
Mature
Aged
II
Mature
Expression
Aged II
Mature
Mature
glutathione
S-transferase 7
XM_013079611.1
potassium
voltage-gated channel protein Shaker-like
XM_005089856.2
fatty
acid-binding protein
XM_005097581.2
glutathione
S-transferase 7
glutathione
S-transferase 7
XM_013079611.1
XM_013079611.1
CCAP-1
XM_005103331.1
potassium
voltage-gated channel protein Shaker-like
aquaporin-5-like
potassium
voltage-gated channel protein Shaker-like
XM_005089856.2
XM_005111003.2
XM_005089856.2
brain-specific
domain protein
3-like
XM_005112117.2 homeobox/POU
fatty
acid-binding protein
fatty
acid-binding
protein
XM_005097581.2
XM_005097581.2
uncharacterized
LOC101862347
XM_013082188.1
CCAP-1
XM_005103331.1
CCAP-1
XM_005103331.1
SKI
family transcriptional
corepressor 1 homolog-B-like
XM_005096911.2
aquaporin-5-like
XM_005111003.2
XM_005111003.2
adenylate
cyclase aquaporin-5-like
NM_001204659.1
brain-specific
3-like
brain-specific
domain protein 3-like
XM_005112117.2 homeobox/POU domain protein glutamate
receptor ionotropic,
kainatehomeobox/POU
2-like
XM_005112117.2
XM_005097743.2
stimulated
by retinoic
acid gene 6 protein LOC101862347
homolog
uncharacterized
LOC101862347
XM_005110741.2
uncharacterized
XM_013082188.1
XM_013082188.1
uncharacterized
LOC101862112
XM_013082201.1
SKI
family transcriptional corepressor 1 homolog-B-like
SKI
family transcriptional corepressor 1 homolog-B-like
XM_005096911.2
XM_005096911.2
glutamate
receptor 2-like
XM_005097744.2
adenylate
cyclase
adenylate
cyclase
NM_001204659.1
NM_001204659.1
homeobox
protein abdominal-A
homolog
XM_005110906.2
glutamate
receptor ionotropic, kainate 2-like uncharacterized
glutamate
receptor ionotropic, kainate 2-like
XM_005097743.2
XM_005097743.2
XM_013082232.1
LOC101851731
stimulated
by retinoic acid gene 6 protein homolog
stimulated
XM_013082230.1
XM_005110741.2
G-protein
coupled receptor
64-likeby retinoic acid gene 6 protein homolog
XM_005110741.2
XM_013082215.1
delta
and Notch-likeuncharacterized
epidermal growth factor-related
receptor
uncharacterized
LOC101862112
LOC101862112
XM_013082201.1
XM_013082201.1
uncharacterized
LOC106011702
XM_013082231.1
glutamate
receptor 2-like
XM_005097744.2
glutamate
receptor 2-like
XM_005097744.2
G2/mitotic-specific
cyclin-B3-like
XM_005100278.2
homeobox
protein abdominal-A homolog
XM_005110906.2
homeobox
protein
abdominal-A
homolog
XM_005110906.2
cAMP-specific
XM_013086739.1 3',5'-cyclic phosphodiesterase 4D-like
XM_013082232.1
uncharacterized
LOC101851731
XM_013082232.1
LOC101851731
XM_013081720.1
protein
enabled-likeuncharacterized
XM_013083348.1
pro-neuregulin-2,
membrane-bound
isoformreceptor 64-like
XM_013082230.1
XM_013082230.1
G-protein
coupled receptor 64-like
G-protein
coupled
XM_005103388.2
dihydroxyacetone
acyltransferase-like
XM_013082215.1
XM_013082215.1
delta
and Notch-like epidermal growth factor-related
receptorphosphate
delta
and
Notch-like epidermal growth factor-related receptor
XM_013080819.1
hemicentin-1
uncharacterized
LOC106011702
uncharacterized
LOC106011702
XM_013082231.1
XM_013082231.1
XM_005103777.2
heat
shock protein 70 B2-like
G2/mitotic-specific
cyclin-B3-like
G2/mitotic-specific
cyclin-B3-like
XM_005100278.2
XM_005100278.2
XM_005108566.2
major
vault protein
XM_005104526.2
cAMP-specific
XM_013086739.1 3',5'-cyclic phosphodiesterase 4D-like
cAMP-specific
4D-like
XM_013086739.1 3',5'-cyclic phosphodiesterase uncharacterized
LOC101864002
XM_005099220.2 calcium
store-operated
entry-associated
regulatory factor
XM_013081720.1
XM_013081720.1
protein
enabled-like
protein
enabled-like
XM_005113313.2
LOC101864506
XM_013083348.1
XM_013083348.1
pro-neuregulin-2,
membrane-bound isoform
pro-neuregulin-2,
membrane-bound isoform uncharacterized
XM_005106432.2
uncharacterized
LOC101846602
XM_005103388.2
XM_005103388.2
dihydroxyacetone
phosphate acyltransferase-like
dihydroxyacetone
phosphate acyltransferase-like
XM_013086971.1
multidrug
resistance-associated
protein 1-like
XM_013080819.1
XM_013080819.1
hemicentin-1
XM_013087976.1
hemicentin-1
elongation factor 2-like
XM_005100777.2
XM_005103777.2
DNA
ligase-like
XM_005103777.2
heat
shock protein 70 B2-like
heat
shock protein 70 B2-like
XM_005104934.2
cat
eye syndrome critical
region
5 homolog
XM_005108566.2
XM_005108566.2
major
vaultprotein
protein
major
vault protein
XM_013085278.1
glutamine
synthetase-like
XM_005104526.2
XM_005104526.2
uncharacterized
LOC101864002
uncharacterized
LOC101864002
XM_005096327.2
myb-like
protein I
XM_005099220.2 calcium entry-associated regulatory factor
XM_005099220.2 calcium entry-associated regulatory
store-operated
store-operated
factor
XM_013089182.1
putative
eggshell protein
XM_005113313.2
XM_013079144.1
XM_005113313.2
uncharacterized
LOC101864506
uncharacterized
LOC101864506
glycine
N-acyltransferase-like
XM_005103166.2
XM_005106432.2
XM_005106432.2
zinc
finger protein 704-like
uncharacterized
LOC101846602
uncharacterized
LOC101846602
XM_005097221.2
receptor-like
XM_013086971.1
XM_013086971.1
multidrug
resistance-associated protein 1-like
multidrug
resistance-associated protein 1-likeFMRFamide
XM_005105439.2
uncharacterized
LOC101855585
XM_013087976.1
XM_013087976.1
elongation
factor 2-like
elongation
factor 2-like
XM_005090908.2
uncharacterized
LOC101861258
XM_005100777.2
XM_005100777.2
DNA
ligase-like
XM_013083177.1
DNA
ligase-like
uncharacterized
LOC101856027
XM_005105964.2
XM_005104934.2
XM_005104934.2
cat
eyedomain-containing
syndrome critical
region
major
facilitator superfamily
protein
1-like protein 5 homolog
cat
eye syndrome critical region protein 5 homolog
XM_013080424.1
XM_013085278.1
uncharacterized
LOC106011289
XM_013085278.1
glutamine
synthetase-like
glutamine
synthetase-like
XM_005093424.2
uncharacterized
LOC101856247
XM_005096327.2
XM_005096327.2
myb-like
protein I
myb-like
protein I
XM_013087517.1
inter-alpha-trypsin
inhibitor
heavy
chain
H4-like
XM_013089182.1
XM_013089182.1
putative
eggshell protein
XM_005104831.2
putative
eggshell protein
uncharacterized
LOC101854594
XM_013083237.1
glycine N-acyltransferase-like
insulin-like
growth factor-binding
protein complex acid labile subunit
XM_005099123.2
XM_005103166.2
M6
M5
M1
M2
M3
M4
M6
A3
M6
M5
A1
M5
M1
A2
M1
M2
A4
M2
M3
M3
A5
M4
XM_005103166.2
XM_005103166.2
uncharacterized LOC101858410
zinc finger protein 704-like
zinc
finger protein 704-like
zinc
finger protein 704-like
XM_005097221.2
XM_005097221.2
XM_005097221.2
FMRFamide
receptor-like
FMRFamide
receptor-like
FMRFamide
receptor-like
XM_005105439.2
XM_005105439.2
XM_005105439.2
uncharacterized LOC101855585
uncharacterized
LOC101855585
uncharacterized LOC101855585
XM_005090908.2
XM_005090908.2
XM_005090908.2
uncharacterized
LOC101861258
uncharacterized
LOC101861258
uncharacterized
LOC101861258
XM_013083177.1
XM_013083177.1
XM_013083177.1
uncharacterized
LOC101856027
uncharacterized
LOC101856027
uncharacterized
LOC101856027
XM_005105964.2
XM_005105964.2
XM_005105964.2
facilitator superfamily domain-contain
major
facilitator superfamily domain-containing protein 1-like major facilitator superfamily domain-containing protein 1-like major
XM_013080424.1
XM_013080424.1
XM_013080424.1
uncharacterized
LOC106011289
uncharacterized
LOC106011289
uncharacterized
LOC106011289
XM_005093424.2
XM_005093424.2
XM_005093424.2
uncharacterized LOC101856247
uncharacterized LOC101856247
uncharacterized LOC101856247
XM_013087517.1
XM_013087517.1
XM_013087517.1
inter-alpha-trypsin
inhibitor heavy chain H4-like
inter-alpha-trypsin
inhibitor heavy chain H4inter-alpha-trypsin
inhibitor heavy chain H4-like
XM_005104831.2
XM_005104831.2
XM_005104831.2
uncharacterized
LOC101854594
uncharacterized
LOC101854594
uncharacterized
LOC101854594
XM_013083237.1
XM_013083237.1
XM_013083237.1
insulin-like
growth factor-binding protein complex acid labile subunit
insulin-like
growth factor-binding protein com
insulin-like
growth factor-binding protein complex acid labile subunit
XM_005099123.2
XM_005099123.2
XM_005099123.2
uncharacterized
LOC101858410
uncharacterized
LOC101858410
uncharacterized
LOC101858410
M4
A6
A3
A3
M6
A2
A1
A1
M5
A4
A2
M1
A5
A4
M2
A6
A5
M3
A6
M4
A3
A1
A2
A4
A5
XM_013079144.1
glycine
N-acyltransferase-like
XM_013079144.1
XM_013079144.1
glycine
N-acyltransferase-like
glutathione
S-transferase 7
XM_013079611.1
potassium
voltage-gated channel protein Sh
XM_005089856.2
fatty
acid-binding protein
XM_005097581.2
CCAP-1
XM_005103331.1
aquaporin-5-like
XM_005111003.2
brain-specific
XM_005112117.2 homeobox/POU domain prote
uncharacterized
LOC101862347
XM_013082188.1
SKI
family transcriptional corepressor 1 hom
XM_005096911.2
adenylate
cyclase
NM_001204659.1
glutamate
receptor ionotropic, kainate 2-like
XM_005097743.2
stimulated
by retinoic acid gene 6 protein ho
XM_005110741.2
uncharacterized
LOC101862112
XM_013082201.1
glutamate
receptor 2-like
XM_005097744.2
homeobox
protein abdominal-A homolog
XM_005110906.2
XM_013082232.1
uncharacterized
LOC101851731
XM_013082230.1
G-protein
coupled receptor 64-like
XM_013082215.1
delta
and Notch-like epidermal growth facto
uncharacterized
LOC106011702
XM_013082231.1
G2/mitotic-specific
cyclin-B3-like
XM_005100278.2
cAMP-specific
XM_013086739.1 3',5'-cyclic phosphodiesteras
XM_013081720.1
protein
enabled-like
XM_013083348.1
pro-neuregulin-2,
membrane-bound isoform
XM_005103388.2
dihydroxyacetone
phosphate acyltransferas
XM_013080819.1
hemicentin-1
XM_005103777.2
heat
shock protein 70 B2-like
XM_005108566.2
major
vault protein
XM_005104526.2
uncharacterized
LOC101864002
XM_005099220.2 calcium entry-associated reg
store-operated
XM_005113313.2
uncharacterized
LOC101864506
XM_005106432.2
uncharacterized
LOC101846602
XM_013086971.1
multidrug
resistance-associated protein 1-lik
XM_013087976.1
elongation
factor 2-like
XM_005100777.2
DNA
ligase-like
XM_005104934.2
cat
eye syndrome critical region protein 5 ho
XM_013085278.1
glutamine
synthetase-like
XM_005096327.2
myb-like
protein I
XM_013089182.1
putative
eggshell protein
Fig 4.5 Heatmap of the top 50 most significantly DE genes from PVC neuron
clusters of M and AII animals. Yellow indicates higher expression compared to the
average across all replicates, and red indicates reduced expression. The role of underlined
genes in the Aplysia nervous system is addressed in results and discussion to illustrate the
aging process. This includes upregulation in AII animals of several stress related genes
such as heat shock protein and major vault protein, and downregulation of many
transcripts related to nervous system function.
117
1.2
Relative expression compared to Mature
1
0.8
*
0.6
*
0.4
*
0.2
NMDA
AMPA
KA2
GluR7
GluR6
GluR8
GluR5
GluR3
GluR2
GluR1
Grin2
Grin1
0
Kainate/orphans
Subunit
Fig. 4.6. Expression of iGluR subunits in M and AII PVC neurons based on
RNASeq. Relative expression of iGluR transcripts in AII compared to M in PVC sensory
neurons showed down-regulation of some iGluR genes in AII. There were no iGluR
subunits with increased expression during aging. KA1 and GluR4 subunits did not meet
the threshold for number of transcripts to be analyzed.
*denotes significantly decreased expression compared to M (p£0.05, Student’s t-test,
FDR corrected)
118
Mature
2000
Aged II
1800
100
1400
90
1200
1000
B
80
70
800
60
600
50
400
40
200
30
0
20
Grin1
Grin2
Subunit
10
0
100
% of total AMPAR expression
1600
% of total NMDA expression
Calculated number of transcripts
A
90
80
70
60
50
40
30
20
10
0
Grin1
Grin2
Subunit
*
*
GluR1 GluR2 GluR3 GluR5 GluR8
Subunit
Fig. 4.7. Composition of NMDAR and AMPAR in M and AII PVC neurons. The
percent contribution of each subunit to total expression of each subtype was calculated.
A) No significant changes in relative expression of NMDAR subunits Grin1 and Grin2 in
AII PVC neurons were observed. B) For AMPAR subunits, GluR1 comprised a
significantly greater percentage of total AMPAR expression in AII PVC neurons. GluR3
made up a significantly smaller percentage of AMPAR expression in AII PVC neurons.
*denotes significantly different percentage of total subtype expression (p£0.05, one-way
ANOVA on a logistic-regression with quasibinomial distribution)
119
Molecular
Functions
Molecular Functions
Biological
Processes
Biological Processes
activation of
Activation of protein
protein kinase A
activity
kinase A activity
Ion channel
ion channel
associated
associated
nucleotide
Nucleotide metabolic
metabolic process
process
Protein
kinase
protein kinase
bindingbinding
nucleoside
Nucleosidephosphate
phosphate
metabolic
metabolic processprocess
Nucleobase-containing
nucleobase−containing
small molecule
small molecule
metabolic
metabolic process
process
Ubiquitin-like
ubiquitin−like
protein ligase
protein
ligase
binding
binding
carbohydrate
Carbohydrate derivative
derivative
metabolic process
metabolic process
Activation of activation of
phospholipase C
activity
phosolipase C activity
1e−01
Adenyl adenyl
ribonucleotide
ribonucleotide
binding
binding
1e−04
1e−07
1e−10
1e−13
p−value (FDR corrected)
p-value (FDR corrected)
1e−02
1e−05
1e−08
1e−11
p−value (FDR corrected)
p-value (FDR corrected)
Fig. 4.8. The pathways and activities most affected by aging. The most significantly
affected GO terms are listed on the y-axis, with FDR-corrected p-values on the x-axis.
The most significantly enriched molecular function is for ion channel associated genes,
suggesting that many ion channels were affected by aging. Activation of protein kinase A
activity implies that second messenger systems may be altered in aging.
120
2
2
0
0
-1
−1
log 2 fold change
1
1
−2
-2
−3
-3
1e−01
1e-01
1e+01
1e+01
1e+03
1e+03
Mean expression
Fig. 4.9. Mean expression vs. log2 fold change for ion channel associated genes.
Points in red indicate ion channel genes that were DE between M and AII animals
(p£0.05, Student’s t-test, FDR corrected). Negative log-fold changes along the y-axis
corresponds to transcripts that showed decreased expression in AII animals. Despite an
approximately equal number of genes up and down-regulated in the full dataset, >80% of
DE ion channel genes exhibited decreased expression in AII.
121
3
3
1
1
0
0
-1
−1
log 2 fold change
2
2
−2
-2
−3
-3
1e−01
1e-01
1e+01
1e+01
1e+03
1e+03
Mean expression
Fig. 4.10. Mean expression vs. log2 fold change for stress response genes. Points in red
indicate stress response genes that were DE between M and AII animals (p£0.05,
Student’s t-test, FDR corrected). Positive log-fold changes along the y-axis corresponds
to transcripts that showed increased expression in AII animals. Many species have shown
increased stress response with age, and >70% of DE stress response genes in Aplysia
were upregulated in AII PVC neurons.
122
1.5
RNAseq
qPCR
*
log 2 fold change
1
0.5
0
-0.5
*
*
*
*
-1
*
*
-1.5
1
2
3
4
5
6
7
8
*
9
Fig 4.11. Comparison of DE genes in RNASeq and qPCR analyses. Expression status
of genes important for neuronal excitability and genes implicated in aging from RNASeq
were further tested via qPCR. Log2 fold change in expression in AII animals is displayed
on the y-axis for both RNAseq and qPCR results. Negative log-fold change indicates
reduced expression in AII animals. K+ Shaw and Ca+ subunit alpha were the only two
genes significantly DE in both analyses (qPCR n=4, RNASeq n=6).
*denotes significantly different expression compared to M
Chapter 5:
Conclusions
Invertebrate species such as Aplysia californica (Aplysia), Drosophila
melanogaster, and Caenorhabditis elegans have been used for decades as models of
nervous system function. This dissertation focused on studies of the receptors for LGlutamate (L-Glu). L-Glu is the primary neurotransmitter for fast synaptic transmission
in nearly all animals. The binding of L-Glu to ionotropic L-Glu receptors (iGluR) on the
postsynaptic membrane opens transmembrane ion channels that allow positively charged
ions to flow into the cell, leading to depolarization that can trigger an action potential and
transmit information. The simplified nervous systems of invertebrates compared to
vertebrates allows for detailed studies of iGluR to further understand processes relevant
in vertebrates, such as humans.
A better understanding of the relationship between iGluR subunits in Bilateria
both clarifies and strengthens the power of animal models of iGluR-mediated
neurotransmission. In Chapter 2, the evolutionary relationship of iGluR subunits was
examined to identify invertebrate iGluR subunits most similar to the three vertebrate
iGluR subtypes: N-methyl-d-aspartate receptors (NMDAR), α-Amino-3-hydroxy-5methyl-4-isoxazolepropionic acid receptors (AMPAR), and kainate receptors.
The strongest conclusion of this analysis was that the NMDAR subtype is the
most well-conserved across bilaterian species analyzed. Despite large evolutionary
distance, both Grin1 and Grin2 NMDAR subunits formed orthologous groups across all
species. This suggests that there is high conservation of individual NMDAR subunits.
High selective constraint implies that NMDAR function has been conserved throughout
123
124
Bilateria, and thus NMDAR function is necessary for proper nervous system
function in all species studied. This is unique to NMDAR; it is the only iGluR subtype
that contained more than one gene prior to the divergence of protostomes and
deuterostomes.
One goal of this research was to predict subunits that form functional iGluR in
Aplysia. In a phylogenetic analysis of vertebrate iGluR subunits, NMDAR, AMPAR, and
kainate receptor subunits each form a monophyletic clade, with functional iGluR formed
as obligatory tetramers composed only of subunits within each clade (Dingledine, et al.
1999). The phylogenetic analysis in Aplysia identified a well-defined clade for the
NMDAR subunits that were supported in the full phylogeny. An Aplysia NMDAR-like
iGluR must employ Grin1-1 or Grin1-2, and Grin2, if similar to the vertebrate
heterotetrameric structure requiring two Grin1 and 2 Grin2 subunits (Paoletti and Neyton
2007). Since expression of the Aplysia Grin1-1 subunit alone did not form functional
iGluR in Xenopus laevis oocytes (Ha, et al. 2006), the NMDAR-like iGluR cannot be a
homotetramer. This suggests that tetrameric Aplysia NMDAR are similar to vertebrate
NMDAR in the requirement of both Grin1 and Grin2 subunits for functional receptors.
All 12 iGluR subunits were expressed in all Aplysia nervous system tissues. Prior
to this dissertation, the breadth and reach of the role of iGluR, and of L-Glu as a
neurotransmitter in Aplysia, was unclear. The ubiquitous expression of numerous iGluR
subunits throughout the nervous system attests to the importance of glutamatergic
neurotransmission in Aplysia, as in mammals.
High conservation of individual NMDAR subunits and preliminary indications
that Aplysia NMDAR may require both Grin1 and Grin2 subunits suggests that many
125
aspects of NMDAR physiology are conserved in bilaterians. This contention supports the
use of invertebrate species as models for studies of NMDA-mediated physiology.
There are other aspects of NMDAR physiology, however, that are likely different
between vertebrates and at least, marine, invertebrates. In vertebrates, glycine is a coagonist that binds Grin1 subunits and is required for opening of NMDAR. Seawater
contains large amounts of glycine (Hubberten, et al. 1994) that make it an unlikely coagonist in Aplysia NMDAR; 1 mm added glycine potentiated presumable iGluR currents
activated by D-Aspartate (D-Asp) only at -30 mV. The finding that hydrophobicity of the
Grin1 glycine binding site is not well-conserved in Aplysia further suggests glycine is
unlikely to act as co-agonist. Further, vertebrate NMDAR are subject to allosteric
modulation by extracellular Mg2+, which blocks channel opening constitutively near the
resting potential (Mayer, et al. 1984), where the driving force is low. Mg2+ only weakly
blocked L-Glu-induced excitatory currents in Aplysia motoneurons (Dale and Kandel
1993), while Carlson and Fieber (2011) found that Mg2+ attenuated D-Asp-activated
currents in buccal sensory neurons equally at both hyperpolarized and depolarized
voltages. These studies suggest that any modulatory role of Mg2+ is reduced or that its
role is different in Aplysia NMDAR. Although other small molecules or ions, such as H+
and Zn2+, can act as allosteric modulators of vertebrate NMDAR (Paoletti 2011), there is
currently no evidence that any of these molecules perform modulatory functions in
NMDAR-like-mediated neurophysiology in Aplysia; however, these are understudied
areas. In summary, some of the actions of vertebrate NMDAR allosteric modulators may
not be relevant in Aplysia.
126
NMDAR in vertebrates are primarily Ca2+ channels, and thus their function is
tightly regulated to control intracellular Ca2+ concentration. Carlson and Fieber (2011)
found that presumable iGluR activated by D-Asp did not conduct Ca2+. If Aplysia
NMDAR are not Ca2+-permeable, it may explain why they are not as tightly controlled
through allosteric modulators. It is also possible that other molecules, not yet identified,
play the modulatory role in Aplysia NMDAR, with Mg2+- and glycine-dependence, as
well as Ca2+ permeability, evolving in vertebrates after divergence from invertebrates.
NMDAR is the only iGluR subtype in vertebrates with these unique characteristics. The
similarity of NMDAR in Aplysia and vertebrates emphasizes the important, ancestral role
for NMDAR that has been maintained through functional constraint throughout Bilateria.
While Grin1 and Grin2 were identified as the likely eponymous NMDAR
subunits in Aplysia, more research is necessary to verify that these mediate observed
NMDAR currents. New genome and transcriptome databases have identified Grin1-2 and
Grin2 NMDAR subunits since the Xenopus expression study of Grin1 described above
(Ha, et al. 2006). Co-expression of vertebrate Grin1 and Grin2 subunits can form
complete functional receptors in Xenopus (Kuner and Schoepfer 1996; Laube, et al.
1998). By injecting various combinations of Aplysia NMDAR subunits, it could be tested
if these elicit inward currents with pharmacological similarity to those recorded in
Aplysia neurons in previous studies (Carlson and Fieber 2012; Kempsell and Fieber
2014). If Grin1 and Grin2 subunits together form functional receptors, it would further
support the high conservation of NMDAR suggested by the full phylogeny.
AMPAR subunits also form an orthologous group, with one AMPAR gene in the
common bilaterian ancestor. There have been multiple rounds of gene duplication of
127
AMPAR genes in Aplysia, which has likely changed some of the receptor structure. This
may explain why AMPA is not an effective agonist in Aplysia neurons (Trudeau and
Castellucci 1993; Carlson and Fieber 2012). Block of synaptic transmission by AMPAR
antagonists CNQX and DNQX suggests that AMPAR are relevant in Aplysia (Chitwood,
et al. 2001). It is possible that the AMPA agonist-specific binding site of Aplysia
AMPAR has been altered compared to vertebrates, causing AMPA to no longer act as an
agonist.
Returning to and expanding the importance of NMDAR, both qPCR and RNASeq
showed Grin1-1 was expressed at significantly greater copies than all other iGluR
subunits. This suggested that NMDAR, and Grin1-1 in particular, mediate glutamatergic
responses in PVC and BSC sensory neurons of Aplysia. Grin1-1 expression was much
higher in PVC and BSC sensory neuron clusters in (Chapter 3) than in whole ganglia
(Chapter 2). In situ hybridization staining for Grin1 in PVC and BSC sensory neurons
was high compared to whole ganglia (Ha, et al. 2006). Moreover, L-Glu–induced
excitability was a hallmark of PVC neurons in culture (Kempsell and Fieber 2014).
Together the results presented in this dissertation support the literature on NMDAR-like
physiology and expression in Aplysia, and they suggest that Grin1-1 mediates the
majority of iGluR activity in sensory neurons, and by extension, that NMDAR-like
receptor channels form the basis of Aplysia iGluR-mediated neurotransmission in these
neurons. In other tissues in which Grin1 expression is not as high, contributions from
other non-NMDAR subunits may contribute more to L-Glu-induced neurotransmission.
While Grin1 may compose iGluR in Aplysia, the characteristics of the currents
are less classically NMDAR, both electrophysiologically and pharmacologically. For
128
example, NMDA-induced currents were rare in BSC sensory neurons, and, when present,
were of smaller density compared to those induced by the agonists D-Asp or L-Glu
(Carlson and Fieber 2012). Large evolutionary distance between Aplysia and vertebrates
makes it unlikely that these receptors would be exactly the same across Bilateria. Binding
sites for the agonists L-Glu and NMDA are distinct and, in vivo, L-Glu is the likely
agonist at NMDAR. Therefore, selective constraint on the NMDA binding site is low
compared to the L-Glu binding site. In Chapter 2, the NMDA binding site was observed
to be less similar in hydrophobicity than transmembrane domains (TMDs) that form the
ion channel. Therefore, NMDA may not effectively bind Aplysia NMDAR.
The expression studies showed no change in subunit composition with age
occurred for either NMDAR or AMPAR, suggesting that reduced L-Glu currents in
sensory neurons of aged animals is caused by a loss of functional iGluR. Several iGluR
subunits had significantly decreased expression in AII animals, and zero subunits had
significantly increased expression via qPCR. Grin1-1, the dominant subunit, showed
reduced expression in aged PVC and BSC neurons of both cohorts studied. This was an
important result. As the subunit that likely mediates most iGluR currents, downregulation
and loss of Grin1-1 receptors is the basis of reduced iGluR-mediated excitability in aged
sensory neurons, if receptor loss equals decline in receptor physiology. Although their
role is not as clear, reduced expression of other iGluR subunits, variable in identity
among cohorts, may contribute to decline in receptor physiology in some way. Altered
composition of functional iGluR, changing the subunits that make up the receptor, has
been implicated as a contributor to reduced iGluR physiology with age in vertebrates
129
(Barnes, et al. 1992; Barnes, et al. 1997; Jouvenceau, et al. 1998), so in this sense Aplysia
aging is unique.
Yet in other ways, loss of Aplysia Grin1-1 with age echoes the pattern in other
aging models. Vertebrate NMDAR may be more vulnerable to aging effects compared to
other iGluR subtypes. Many layers of the mouse brain showed greater declines in NMDA
binding than AMPA binding in aged animals (Magnusson and Cotman 1993; Magnusson
1995, 1997). Similar results have also been observed in dogs, monkeys, and rats (Tamaru,
et al. 1991; Magnusson 2000; Hof, et al. 2002).
The whole-transcriptome changes in aging Aplysia sensory neurons suggest the
critical loss of iGluR is only part of the aging story. Genes associated with baseline
neuronal excitability were significantly affected. The most striking observation was that
>80% of the ion channel genes that were differentially expressed had reduced expression,
with many coding for voltage-gated K+ and Ca2+ channels. The reduced baseline
excitability that would result likely contributes to aging declines in the function of these
sensory neurons. Interestingly, gene expression may be a sensitive index of aging
changes in excitability, since no changes in Ca2+ currents (Fieber, et al. 2010), or in
depolarization amplitude, afterhyperpolarization, or duration of AP (Kempsell and Fieber
2015b) were found in aged Aplysia sensory neurons.
Reduced neuronal excitability as a hallmark of aging in the mammalian
hippocampus is correlated with memory declines (Barnes 1990; Barnes, et al. 1992;
Foster and Norris 1997; Jouvenceau, et al. 1998; Kumar and Foster 2007). Reduced
intrinsic excitability in hippocampal pyramidal neurons occurred by alterations in
afterhyperpolarization (AHP), which is primarily mediated by voltage-gated K+ channels
130
(Disterhoft and Oh 2007). Thus, K+ channel attrition may be a factor in memory and
cognitive declines, and also may be a conserved process that occurs in all animals,
including Aplysia. Here is another area of research in which the identified contributions
of Aplysia neurons to specific reflex behaviors can be leveraged as a model.
A ubiquitous mechanism of aging in vertebrate models is DNA damage caused by
excessive oxygen radicals. These reactive oxygen species (ROS) accumulate in tissues of
old animals, causing oxidative damage. Oxidation of ion channels is prevalent in aging
vertebrate brains and can alter functional properties of voltage-gated Na+, K+, and Ca2+
channels (Annunziato, et al. 2002; Patel and Sesti 2016). The cellular response to
mitigate the effects of ROS is increased transcription of stress response genes. Stress
response genes can refold or degrade proteins that have been damaged by ROS.
In aged PVC sensory neurons there was increased transcription of many stress
response genes. Several that were upregulated are conserved between vertebrates and
Aplysia, including heat shock proteins, major vault protein, and multi-drug resistance
protein. Increased stress response has been correlated with increased ROS production.
Thus, increased ROS in aged Aplysia, resulting in increased stress response gene
transcription, may be a major component of aging effects in Aplysia, as it is in vertebrate
models.
Neurodegenerative diseases are a common cause of reduced physiology and
cognitive declines in old humans. For example, abnormal protein accumulation that
results in reduced neural function has been linked to diseases such as Alzheimer’s,
Parkinson’s, and Huntington’s. Altered transcription specific genes, such as amyloid
proteins and huntingtin, have been correlated with several of these diseases. There was no
131
change in expression for amyloid proteins or huntingtin in aged PVC neurons. These
observations, when coupled to the reduced neurophysiological function observed in
Aplysia near the end of its annual lifespan, provide some evidence that Aplysia models
the normal aging process, rather than neurological disease.
The aging process results in large scale gene expression changes that could be
responsible for significant changes in neural function in aged animals. Many of these
pathways that occur in vertebrates, such as reduced neural excitability, reduced response
to neurotransmitters such as L-Glu, and increases in stress response gene transcription,
were found to also occur in Aplysia. The findings of this dissertation suggest that many
aspects of aging in the nervous system are conserved in Aplysia and vertebrates. Thus,
Aplysia provide a relevant model system for the effects of aging with relevance to the
vertebrate nervous system and is promising for the continued use of Aplysia for the
neurobiology of aging with relevance to vertebrates.
References
Akazawa C, Shigemoto R, Bessho Y, Nakanishi S, Mizuno N. 1994. Differential
expression of five N-methyl-D-aspartate receptor subunit mRNAs in the
cerebellum of developing and adult rats. Journal of Comparative Neurology
347:150-160.
Akhmedov K, Rizzo V, Kadakkuzha BM, Carter CJ, Magoski NS, Capo TR,
Puthanveettil SV. 2013. Decreased response to acetylcholine during aging of
aplysia neuron R15. PloS one 8:e84793.
Alberini CM, Ghirardl M, Metz R, Kandel ER. 1994. C/EBP is an immediate-early gene
required for the consolidation of long-term facilitation in Aplysia. Cell 76:10991114.
Alivisatos AP, Chun M, Church GM, Greenspan RJ, Roukes ML, Yuste R. 2012. The
brain activity map project and the challenge of functional connectomics. Neuron
74:970-974.
Anders S, Pyl PT, Huber W. 2014. HTSeq—a Python framework to work with highthroughput sequencing data. Bioinformatics:btu638.
Annunziato L, Pannaccione A, Cataldi M, Secondo A, Castaldo P, Di Renzo G,
Taglialatela M. 2002. Modulation of ion channels by reactive oxygen and
nitrogen species: a pathophysiological role in brain aging? Neurobiology of aging
23:819-834.
Bacskai BJ, Hochner B, Mahaut-Smith M, Adams SR, Kaang B-K, Kandel ER, Tsien
RY. 1993. Spatially resolved dynamics of cAMP and protein kinase A subunits in
Aplysia sensory neurons. Science 260:222-222.
Bailey CH, Bartsch D, Kandel ER. 1996. Toward a molecular definition of long-term
memory storage. Proceedings of the National Academy of Sciences 93:1344513452.
Bailey CH, Castellucci VF, Koester J, Chen M. 1983. Behavioral changes in aging
Aplysia: a model system for studying the cellular basis of age-impaired learning,
memory, and arousal. Behavioral and neural biology 38:70-81.
Bailey CH, Chen M. 1988a. Long-term memory in Aplysia modulates the total number of
varicosities of single identified sensory neurons. Proceedings of the National
Academy of Sciences 85:2373-2377.
Bailey CH, Chen M. 1988b. Long-term sensitization in Aplysia increases the number of
presynaptic contacts onto the identified gill motor neuron L7. Proceedings of the
National Academy of Sciences 85:9356-9359.
132
133
Barnes CA. 1990. Effects of aging on the dynamics of information processing and
synaptic weight changes in the mammalian hippocampus. Progress in brain
research 86:89-104.
Barnes CA, Rao G, Foster T, McNaughton B. 1992. Region-specific age effects on
AMPA sensitivity: Electrophysiological evidence for loss of synaptic contacts in
hippocampal field CA1. Hippocampus 2:457-468.
Barnes CA, Rao G, Shen J. 1997. Age-related decrease in the N-methyl-D-aspartate Rmediated excitatory postsynaptic potential in hippocampal region CA1.
Neurobiology of aging 18:445-452.
Bartsch D, Ghirardi M, Skehel PA, Karl KA, Herder SP, Chen M, Bailey CH, Kandel
ER. 1995. Aplysia CREB2 represses long-term facilitation: relief of repression
converts transient facilitation into long-term functional and structural change. Cell
83:979-992.
Bishop NA, Lu T, Yankner BA. 2010. Neural mechanisms of ageing and cognitive
decline. Nature 464:529-535.
Blankenberg D, Gordon A, Von Kuster G, Coraor N, Taylor J, Nekrutenko A. 2010.
Manipulation of FASTQ data with Galaxy. Bioinformatics 26:1783-1785.
Bordner KA, Kitchen RR, Carlyle B, George ED, Mahajan MC, Mane SM, Taylor JR,
Simen AA. 2011. Parallel declines in cognition, motivation, and locomotion in
aging mice: association with immune gene upregulation in the medial prefrontal
cortex. Experimental gerontology 46:643-659.
Boxer PA, Morales FR, Chase MH. 1988. Alterations of group Ia-motoneuron
monosynaptic EPSPs in aged cats. Experimental neurology 100:583-595.
Bradford J, Shin J-Y, Roberts M, Wang C-E, Li X-J, Li S. 2009. Expression of mutant
huntingtin in mouse brain astrocytes causes age-dependent neurological
symptoms. Proceedings of the National Academy of Sciences 106:22480-22485.
Braunersreuther V, Jaquet V. 2012. Reactive oxygen species in myocardial reperfusion
injury: from physiopathology to therapeutic approaches. Current pharmaceutical
biotechnology 13:97-114.
Brockie PJ, Madsen DM, Zheng Y, Mellem J, Maricq AV. 2001. Differential expression
of glutamate receptor subunits in the nervous system of Caenorhabditis elegans
and their regulation by the homeodomain protein UNC-42. The Journal of
Neuroscience 21:1510-1522.
Brunelli M, Castellucci V, Kandel E. 1976. Synaptic facilitation and behavioral
sensitization in Aplysia: possible role of serotonin and cyclic AMP. Science
194:1178-1181.
134
Byrne JH, Castellucci VF, Carew TJ, Kandel ER. 1978. Stimulus-response relations and
stability of mechanoreceptor and motor neurons mediating defensive gillwithdrawal reflex in Aplysia. Journal of Neurophysiology 41:402-417.
Byrne JH, Castellucci VF, Kandel ER. 1978. Contribution of individual mechanoreceptor
sensory neurons to defensive gill-withdrawal reflex in Aplysia. Journal of
Neurophysiology 41:418-431.
Cai D, Chen S, Glanzman DL. 2008. Postsynaptic regulation of long-term facilitation in
Aplysia. Current Biology 18:920-925.
Calderwood SK, Murshid A, Prince T. 2009. The shock of aging: molecular chaperones
and the heat shock response in longevity and aging–a mini-review. Gerontology
55:550-558.
Capella-Gutiérrez S, Silla-Martínez JM, Gabaldón T. 2009. trimAl: a tool for automated
alignment trimming in large-scale phylogenetic analyses. Bioinformatics 25:19721973.
Capo TR, Fieber LA, Stommes DL, Walsh PJ. 2003. Reproductive output in the
hatchery-reared California sea hare at different stocking densities. Journal of the
American Association for Laboratory Animal Science 42:31-35.
Carew TJ, Castellucci VF, Kandel ER. 1971. An analysis of dishabituation and
sensitization of the gill-withdrawal reflex in Aplysia. International Journal of
Neuroscience 2:79-98.
Carew TJ, Walters ET, Kandel ER. 1981. Classical conditioning in a simple withdrawal
reflex in Aplysia californica. The Journal of Neuroscience 1:1426-1437.
Carlson SL, Fieber LA. 2011. Physiological evidence that D-aspartate activates a current
distinct from ionotropic glutamate receptor currents in Aplysia californica
neurons. Journal of Neurophysiology 106:1629-1636.
Carlson SL, Fieber LA. 2012. Unique ionotropic receptors for D-aspartate are a target for
serotonin-induced synaptic plasticity in Aplysia californica. Comparative
Biochemistry and Physiology Part C: Toxicology & Pharmacology 155:151-159.
Carlson SL, Kempsell AT, Fieber LA. 2012. Pharmacological evidence that D-aspartate
activates a current distinct from ionotropic glutamate receptor currents in Aplysia
californica. Brain Behav 2:391-401.
Castellano JM, Kim J, Stewart FR, Jiang H, DeMattos RB, Patterson BW, Fagan AM,
Morris JC, Mawuenyega KG, Cruchaga C. 2011. Human apoE isoforms
differentially regulate brain amyloid-β peptide clearance. Science translational
medicine 3:89ra57.
135
Castellucci V, Blumenfeld H, Goelet P, Kandel E. 1989. Inhibitor of protein synthesis
blocks longterm behavioral sensitization in the isolated gill-withdrawal reflex of
Aplysia. Journal of neurobiology 20:1-9.
Castellucci VF, Kandel ER, Schwartz JH, Wilson FD, Nairn AC, Greengard P. 1980.
Intracellular injection of t he catalytic subunit of cyclic AMP-dependent protein
kinase simulates facilitation of transmitter release underlying behavioral
sensitization in Aplysia. Proceedings of the National Academy of Sciences
77:7492-7496.
Cedar H, Schwartz JH. 1972. Cyclic adenosine monophosphate in the nervous system of
Aplysia californica II. Effect of serotonin and dopamine. The Journal of general
physiology 60:570-587.
Chalfie M, Sulston JE, White JG, Southgate E, Thomson JN, Brenner S. 1985. The neural
circuit for touch sensitivity in Caenorhabditis elegans. The Journal of
Neuroscience 5:956-964.
Chitwood RA, Li Q, Glanzman DL. 2001. Serotonin facilitates AMPA-type responses in
isolated siphon motor neurons of Aplysia in culture. The Journal of physiology
534:501-510.
Choi Y-B, Kadakkuzha BM, Liu X-A, Akhmedov K, Kandel ER, Puthanveettil SV.
2014. Huntingtin Is Critical Both Pre-and Postsynaptically for Long-Term
Learning-Related Synaptic Plasticity in Aplysia. PloS one 9:e103004.
Chung SG, van Rey EM, Bai Z, Rogers MW, Roth EJ, Zhang L-Q. 2005. Aging-related
neuromuscular changes characterized by tendon reflex system properties.
Archives of physical medicine and rehabilitation 86:318-327.
Circu ML, Aw TY. 2010. Reactive oxygen species, cellular redox systems, and apoptosis.
Free Radical Biology and Medicine 48:749-762.
Coffin AB, Mohr RA, Sisneros JA. 2012. Saccular-specific hair cell addition correlates
with reproductive state-dependent changes in the auditory saccular sensitivity of a
vocal fish. The Journal of Neuroscience 32:1366-1376.
Collingridge G, Herron C, Lester R. 1988. Synaptic activation of N-methyl-D-aspartate
receptors in the Schaffer collateral-commissural pathway of rat hippocampus. The
Journal of physiology 399:283-300.
Contractor A, Swanson GT, Sailer A, O'Gorman S, Heinemann SF. 2000. Identification
of the kainate receptor subunits underlying modulation of excitatory synaptic
transmission in the CA3 region of the hippocampus. The Journal of Neuroscience
20:8269-8278.
Cull-Candy SG, Leszkiewicz DN. 2004. Role of distinct NMDA receptor subtypes at
central synapses. sci STKE 2004:re16.
136
Dai J, Zhou H-X. 2013. An NMDA receptor gating mechanism developed from MD
simulations reveals molecular details underlying subunit-specific contributions.
Biophysical journal 104:2170-2181.
Dale N, Kandel ER. 1993. L-glutamate may be the fast excitatory transmitter of Aplysia
sensory neurons. Proceedings of the National Academy of Sciences 90:71637167.
Danysz W, Zajaczkowski W, Parsons C. 1995. Modulation of learning processes by
ionotropic glutamate receptor ligands. Behavioural pharmacology 6:455-474.
De Loof H, Rosseneu M, Brasseur R, Ruysschaert JM. 1986. Use of hydrophobicity
profiles to predict receptor binding domains on apolipoprotein E and the low
density lipoprotein apolipoprotein BE receptor. Proceedings of the National
Academy of Sciences 83:2295-2299.
Dehal P, Boore JL. 2005. Two rounds of whole genome duplication in the ancestral
vertebrate. PLoS biology 3:1700.
Deupree DL, Bradley J, Turner DA. 1993. Age-related alterations in potentiation in the
CA1 region in F344 rats. Neurobiology of aging 14:249-258.
Dingledine R, Borges K, Bowie D, Traynelis SF. 1999. The glutamate receptor ion
channels. Pharmacological reviews 51:7-62.
Disterhoft JF, Oh MM. 2007. Alterations in intrinsic neuronal excitability during normal
aging. Aging cell 6:327-336.
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M,
Gingeras TR. 2013. STAR: ultrafast universal RNA-seq aligner. Bioinformatics
29:15-21.
Drake TJ, Jezzini S, Lovell P, Moroz LL, Tan W. 2005. Single cell glutamate analysis in
Aplysia sensory neurons. Journal of neuroscience methods 144:73-77.
Dumitriu D, Hao J, Hara Y, Kaufmann J, Janssen WG, Lou W, Rapp PR, Morrison JH.
2010. Selective changes in thin spine density and morphology in monkey
prefrontal cortex correlate with aging-related cognitive impairment. The Journal
of Neuroscience 30:7507-7515.
Dunn CW, Hejnol A, Matus DQ, Pang K, Browne WE, Smith SA, Seaver E, Rouse GW,
Obst M, Edgecombe GD. 2008. Broad phylogenomic sampling improves
resolution of the animal tree of life. Nature 452:745-749.
Edgar RC. 2004. MUSCLE: multiple sequence alignment with high accuracy and high
throughput. Nucleic Acids Res 32:1792-1797.
137
Eldakak A, Rancati G, Rubinstein B, Paul P, Conaway V, Li R. 2010. Asymmetrically
inherited multidrug resistance transporters are recessive determinants in cellular
replicative ageing. Nature cell biology 12:799-805.
Esdin J, Pearce K, Glanzman DL. 2010. Long-term habituation of the gill-withdrawal
reflex in aplysia requires gene transcription, calcineurin and L-type voltage-gated
calcium channels. Front in behavioral neuroscience 4:109-117.
Fayyazuddin A, Villarroel A, Le Goff A, Lerma J, Neyton J. 2000. Four residues of the
extracellular N-terminal domain of the NR2A subunit control high-affinity Zn 2+
binding to NMDA receptors. Neuron 25:683-694.
Fieber LA, Carlson SL, Capo TR, Schmale MC. 2010. Changes in D-aspartate ion
currents in the Aplysia nervous system with aging. Brain research 1343:28-36.
Fieber LA, Carlson SL, Kempsell AT, Greer JB, Schmale MC. 2013. Isolation of sensory
neurons of Aplysia californica for patch clamp recordings of glutamatergic
currents. JoVE (Journal of Visualized Experiments):e50543.
Floyd RA, Hensley K. 2002. Oxidative stress in brain aging: implications for therapeutics
of neurodegenerative diseases. Neurobiology of aging 23:795-807.
Force A, Lynch M, Pickett FB, Amores A, Yan Y-l, Postlethwait J. 1999. Preservation of
duplicate genes by complementary, degenerative mutations. Genetics 151:15311545.
Foster TC, Norris CM. 1997. Age-associated changes in Ca2+-dependent processes:
Relation to hippocampal synaptic plasticity. Hippocampus 7:602-612.
Fox LE, Lloyd PE. 1999. Glutamate is a fast excitatory transmitter at some buccal
neuromuscular synapses in Aplysia. Journal of neurophysiology 82:1477-1488.
Frazier WT, Kandel ER, Kupfermann I, Waziri R, Coggeshall RE. 1967. Morphological
and functional properties of identified neurons in the abdominal ganglion of
Aplysia californica. Journal of Neurophysiology 30:1288-1351.
Gazzaley A, Siegel S, Kordower J, Mufson E, Morrison J. 1996. Circuit-specific
alterations of N-methyl-D-aspartate receptor subunit 1 in the dentate gyrus of
aged monkeys. Proceedings of the National Academy of Sciences 93:3121-3125.
Gerdes R, Fieber LA. 2006. Life history and aging of captive-reared California sea hares
(Aplysia californica). Journal of the American Association for Laboratory Animal
Science 45:40-47.
Gielen M, Retchless BS, Mony L, Johnson JW, Paoletti P. 2009. Mechanism of
differential control of NMDA receptor activity by NR2 subunits. Nature 459:703707.
138
Glanzman DL, Kandel ER, Schacher S. 1990. Target-dependent structural changes
accompanying long-term synaptic facilitation in Aplysia neurons. Science
249:799-802.
Glanzman DL, Mackey SL, Hawkins RD, Dyke AM, Lloyd PE, Kandel ER. 1989.
Depletion of serotonin in the nervous system of Aplysia reduces the behavioral
enhancement of gill withdrawal as well as the heterosynaptic facilitation produced
by tail shock. The Journal of Neuroscience 9:4200-4213.
Goebel DJ, Poosch MS. 1999. NMDA receptor subunit gene expression in the rat brain: a
quantitative analysis of endogenous mRNA levels of NR1 com, NR2A, NR2B,
NR2C, NR2D and NR3A. Molecular Brain Research 69:164-170.
Götz S, García-Gómez JM, Terol J, Williams TD, Nagaraj SH, Nueda MJ, Robles M,
Talón M, Dopazo J, Conesa A. 2008. High-throughput functional annotation and
data mining with the Blast2GO suite. Nucleic acids research 36:3420-3435.
Greger IH, Ziff EB, Penn AC. 2007. Molecular determinants of AMPA receptor subunit
assembly. Trends in neurosciences 30:407-416.
Gundersen V, Fonnum F, Ottersen OP, Storm-Mathisen J. 2001. Redistribution of
Neuroactive Amino Acids in Hippocampus and Striatum During
Hypoglycemia&colon; A Quantitative Immunogold Study. Journal of Cerebral
Blood Flow & Metabolism 21:41-51.
Ha TJ, Kohn AB, Bobkova YV, Moroz LL. 2006. Molecular characterization of NMDAlike receptors in Aplysia and Lymnaea: relevance to memory mechanisms. The
Biological Bulletin 210:255-270.
Halanych KM. 2004. The new view of animal phylogeny. Annual Review of Ecology,
Evolution, and Systematics:229-256.
Hansen KB, Yuan H, Traynelis SF. 2007. Structural aspects of AMPA receptor
activation, desensitization and deactivation. Current opinion in neurobiology
17:281-288.
Hardy J, Selkoe DJ. 2002. The amyloid hypothesis of Alzheimer's disease: progress and
problems on the road to therapeutics. Science 297:353-356.
Hart AK, Fioravante D, Liu R-Y, Phares GA, Cleary LJ, Byrne JH. 2011. Serotoninmediated synapsin expression is necessary for long-term facilitation of the
Aplysia sensorimotor synapse. The Journal of Neuroscience 31:18401-18411.
Herculano-Houzel S. 2009. The human brain in numbers: a linearly scaled-up primate
brain. Frontiers in human neuroscience 3:31.
139
Hirsch HR, Peretz B. 1984. Survival and aging of a small laboratory population of a
marine mollusc, Aplysia californica. Mechanisms of ageing and development
27:43-62.
Hof PR, Duan H, Page TL, Einstein M, Wicinski B, He Y, Erwin JM, Morrison JH. 2002.
Age-related changes in GluR2 and NMDAR1 glutamate receptor subunit protein
immunoreactivity in corticocortically projecting neurons in macaque and patas
monkeys. Brain research 928:175-186.
Hof PR, Morrison JH. 2004. The aging brain: morphomolecular senescence of cortical
circuits. Trends in neurosciences 27:607-613.
Hu J-Y, Glickman L, Wu F, Schacher S. 2004. Serotonin regulates the secretion and
autocrine action of a neuropeptide to activate MAPK required for long-term
facilitation in Aplysia. Neuron 43:373-385.
Hubberten U, Lara R, Kattner G. 1994. Amino acid composition of seawater and
dissolved humic substances in the Greenland Sea. Marine Chemistry 45:121-128.
Hunt DL, Castillo PE. 2012. Synaptic plasticity of NMDA receptors: mechanisms and
functional implications. Current opinion in neurobiology 22:496-508.
Hunter S, Jones P, Mitchell A, Apweiler R, Attwood TK, Bateman A, Bernard T, Binns
D, Bork P, Burge S. 2011. InterPro in 2011: new developments in the family and
domain prediction database. Nucleic acids research 40:D306-D312.
Jouvenceau A, Dutar P, Billard J. 1998. Alteration of NMDA receptor-mediated synaptic
responses in CA1 area of the aged rat hippocampus: Contribution of GABAergic
and cholinergic deficits. Hippocampus 8:627-637.
Kadakkuzha BM, Akhmedov K, Capo TR, Carvalloza AC, Fallahi M, Puthanveettil SV.
2013. Age-associated bidirectional modulation of gene expression in single
identified R15 neuron of Aplysia. BMC genomics 14:880.
Kanda K, Hashizume K, Nomoto E, Asaki S. 1986. The effects of aging on physiological
properties of fast and slow twitch motor units in the rat gastrocnemius.
Neuroscience research 3:242-246.
Kandel ER. 2001. The molecular biology of memory storage: a dialogue between genes
and synapses. Science 294:1030-1038.
Kempsell AT, Fieber LA. 2015a. Age-related deficits in synaptic plasticity rescued by
activating PKA or PKC in sensory neurons of Aplysia californica. Frontiers in
aging neuroscience 7:173.
Kempsell AT, Fieber LA. 2015b. Aging in sensory and motor neurons results in learning
failure in Aplysia californica. PloS one 10:e0127056.
140
Kempsell AT, Fieber LA. 2014. Behavioral aging is associated with reduced sensory
neuron excitability in Aplysia californica. Frontiers in aging neuroscience 6:84.
Kew JN, Kemp JA. 2005. Ionotropic and metabotropic glutamate receptor structure and
pharmacology. Psychopharmacology 179:4-29.
Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. 2013. TopHat2:
accurate alignment of transcriptomes in the presence of insertions, deletions and
gene fusions. Genome Biol 14:R36.
Kimura S, Kawasaki S, Takashima K, Sasaki K. 2001. Physiological and
Pharmacological Characteristics of Quisqualic Acid-Induced K+-Current
Response in the Ganglion Cells of Aplysia. The Japanese journal of physiology
51:511-521.
Klein M, Kandel ER. 1980. Mechanism of calcium current modulation underlying
presynaptic facilitation and behavioral sensitization in Aplysia. Proceedings of the
National Academy of Sciences 77:6912-6916.
Kohda K, Wang Y, Yuzaki M. 2000. Mutation of a glutamate receptor motif reveals its
role in gating and δ2 receptor channel properties. Nature neuroscience 3:315-322.
Kosinski R, Zaremba M. 2007. Dynamics of the model of the Caenorhabditis Elegans
neural network. Acta Physica Polonica B 38:2201-2210.
Kumar A, Foster TC. 2007. 10 Neurophysiology of Old Neurons and Synapses. Brain
aging: models, methods, and mechanisms:229-250.
Kumar S, Stecher G, Tamura K. 2016. MEGA7: Molecular Evolutionary Genetics
Analysis version 7.0 for bigger datasets. Molecular biology and evolution
33:1870-1874.
Kuner T, Schoepfer R. 1996. Multiple structural elements determine subunit specificity
of Mg2+ block in NMDA receptor channels. The Journal of Neuroscience
16:3549-3558.
Landfield PW, Pitler TA, Applegate MD. 1986. The effects of high Mg2+-to-Ca2+ ratios
on frequency potentiation in hippocampal slices of young and aged rats. Journal
of Neurophysiology 56:797-811.
Landis GN, Abdueva D, Skvortsov D, Yang J, Rabin BE, Carrick J, Tavaré S, Tower J.
2004. Similar gene expression patterns characterize aging and oxidative stress in
Drosophila melanogaster. Proceedings of the National Academy of Sciences of
the United States of America 101:7663-7668.
Langmead B, Salzberg SL. 2012. Fast gapped-read alignment with Bowtie 2. Nature
methods 9:357-359.
141
Laube B, Kuhse J, Betz H. 1998. Evidence for a tetrameric structure of recombinant
NMDA receptors. The Journal of Neuroscience 18:2954-2961.
Laurie DJ, Seeburg PH. 1994. Regional and developmental heterogeneity in splicing of
the rat brain NMDAR1 mRNA. The Journal of Neuroscience 14:3180-3194.
Le SQ, Gascuel O. 2008. An improved general amino acid replacement matrix.
Molecular biology and evolution 25:1307-1320.
Leak RK. 2014. Heat shock proteins in neurodegenerative disorders and aging. Journal of
cell communication and signaling 8:293-310.
Lee C-K, Weindruch R, Prolla TA. 2000. Gene-expression profile of the ageing brain in
mice. Nature genetics 25:294-297.
Lee H-K, Takamiya K, Han J-S, Man H, Kim C-H, Rumbaugh G, Yu S, Ding L, He C,
Petralia RS. 2003. Phosphorylation of the AMPA receptor GluR1 subunit is
required for synaptic plasticity and retention of spatial memory. Cell 112:631643.
Letunic I, Doerks T, Bork P. 2012. SMART 7: recent updates to the protein domain
annotation resource. Nucleic acids research 40:D302-D305.
Letunic I, Doerks T, Bork P. 2015. SMART: recent updates, new developments and
status in 2015. Nucleic Acids Res 43:D257-D260.
Levenson J, Endo S, Kategaya LS, Fernandez RI, Brabham DG, Chin J, Byrne JH, Eskin
A. 2000. Long-term regulation of neuronal high-affinity glutamate and glutamine
uptake in Aplysia. Proceedings of the National Academy of Sciences 97:1285812863.
Levenson J, Sherry DM, Dryer L, Chin J, Byrne JH, Eskin A. 2000. Localization of
glutamate and glutamate transporters in the sensory neurons of Aplysia. Journal of
Comparative Neurology 423:121-131.
Li Q, Roberts AC, Glanzman DL. 2005. Synaptic facilitation and behavioral
dishabituation in Aplysia: dependence on release of Ca2+ from postsynaptic
intracellular stores, postsynaptic exocytosis, and modulation of postsynaptic
AMPA receptor efficacy. The Journal of neuroscience 25:5623-5637.
Liao Y, Smyth GK, Shi W. 2014. featureCounts: an efficient general purpose program for
assigning sequence reads to genomic features. Bioinformatics 30:923-930.
Lipsky RH, Goldman D. 2003. Genomics and Variation of Ionotropic Glutamate
Receptors. Annals of the New York Academy of Sciences 1003:22-35.
Lisman JE, Fellous J-M, Wang X-J. 1998. A role for NMDA-receptor channels in
working memory. Nature neuroscience 1:273-275.
142
Liu R-Y, Shah S, Cleary LJ, Byrne JH. 2011. Serotonin-and training-induced dynamic
regulation of CREB2 in Aplysia. Learning & memory 18:245-249.
Loerch PM, Lu T, Dakin KA, Vann JM, Isaacs A, Geula C, Wang J, Pan Y, Gabuzda
DH, Li C. 2008. Evolution of the aging brain transcriptome and synaptic
regulation. PloS one 3:e3329.
Love MI, Huber W, Anders S. 2014. Moderated estimation of fold change and dispersion
for RNA-seq data with DESeq2. Genome Biol 15:550.
Low C-M, Lyuboslavsky P, French A, Le P, Wyatte K, Thiel WH, Marchan EM, Igarashi
K, Kashiwagi K, Gernert K. 2003. Molecular determinants of proton-sensitive Nmethyl-D-aspartate receptor gating. Molecular pharmacology 63:1212-1222.
Lu T, Pan Y, Kao S-Y, Li C, Kohane I, Chan J, Yankner BA. 2004. Gene regulation and
DNA damage in the ageing human brain. Nature 429:883-891.
Luebke J, Chang Y-M, Moore T, Rosene D. 2004. Normal aging results in decreased
synaptic excitation and increased synaptic inhibition of layer 2/3 pyramidal cells
in the monkey prefrontal cortex. Neuroscience 125:277-288.
Luine V, Bowling D, Hearns M. 1990. Spatial memory deficits in aged rats: contributions
of monoaminergic systems. Brain research 537:271-278.
Mackey S, Kandel E, Hawkins R. 1989. Identified serotonergic neurons LCB1 and RCB1
in the cerebral ganglia of Aplysia produce presynaptic facilitation of siphon
sensory neurons. The Journal of Neuroscience 9:4227-4235.
Magnusson KR. 2000. Declines in mRNA expression of different subunits may account
for differential effects of aging on agonist and antagonist binding to the NMDA
receptor. The Journal of Neuroscience 20:1666-1674.
Magnusson KR. 1995. Differential effects of aging on binding sites of the activated
NMDA receptor complex in mice. Mechanisms of ageing and development
84:227-243.
Magnusson KR. 1997. Influence of dietary restriction on ionotropic glutamate receptors
during aging in C57B1 mice. Mechanisms of ageing and development 95:187202.
Magnusson KR, Cotman CW. 1993. Age-related changes in excitatory amino acid
receptors in two mouse strains. Neurobiology of aging 14:197-206.
Magnusson KR, Nelson SE, Young AB. 2002. Age-related changes in the protein
expression of subunits of the NMDA receptor. Molecular Brain Research 99:4045.
143
Malenka RC, Bear MF. 2004. LTP and LTD: an embarrassment of riches. Neuron 44:521.
Mansour M, Nagarajan N, Nehring RB, Clements JD, Rosenmund C. 2001. Heteromeric
AMPA receptors assemble with a preferred subunit stoichiometry and spatial
arrangement. Neuron 32:841-853.
Manzon LA. 2002. The role of prolactin in fish osmoregulation: a review. General and
comparative endocrinology 125:291-310.
Marrus SB, Portman SL, Allen MJ, Moffat KG, DiAntonio A. 2004. Differential
localization of glutamate receptor subunits at the Drosophila neuromuscular
junction. The Journal of Neuroscience 24:1406-1415.
Mayer M, Westbrook G. 1987. Permeation and block of N-methyl-D-aspartic acid
receptor channels by divalent cations in mouse cultured central neurones. The
Journal of physiology 394:501.
Mayer ML, Westbrook GL, Guthrie PB. 1984. Voltage-dependent block by Mg2+ of
NMDA responses in spinal cord neurones. Nature 309:261-263.
McKibben JR, Bass AH. 1998. Behavioral assessment of acoustic parameters relevant to
signal recognition and preference in a vocal fish. The Journal of the Acoustical
Society of America 104:3520-3533.
Meldrum BS. 2000. Glutamate as a neurotransmitter in the brain: review of physiology
and pathology. The Journal of nutrition 130:1007S-1015S.
Mitchell A, Chang H-Y, Daugherty L, Fraser M, Hunter S, Lopez R, McAnulla C,
McMenamin C, Nuka G, Pesseat S. 2014. The InterPro protein families database:
the classification resource after 15 years. Nucleic acids research 43:213-221.
Mitchell JJ, Anderson KJ. 1998. Age-related changes in [3 H] MK-801 binding in the
Fischer 344 rat brain. Neurobiology of aging 19:259-265.
Mohn AR, Gainetdinov RR, Caron MG, Koller BH. 1999. Mice with reduced NMDA
receptor expression display behaviors related to schizophrenia. Cell 98:427-436.
Moroz LL, Edwards JR, Puthanveettil SV, Kohn AB, Ha T, Heyland A, Knudsen B,
Sahni A, Yu F, Liu L. 2006. Neuronal transcriptome of Aplysia: neuronal
compartments and circuitry. Cell 127:1453-1467.
Moroz LL, Kocot KM, Citarella MR, Dosung S, Norekian TP, Povolotskaya IS,
Grigorenko AP, Dailey C, Berezikov E, Buckley KM, et al. 2014. The ctenophore
genome and the evolutionary origins of neural systems. Nature 510:109-114.
144
Moroz LL, Kohn AB. 2010. Do different neurons age differently? Direct genome-wide
analysis of aging in single identified cholinergic neurons. Frontiers in aging
neuroscience 2:1-18.
Murshid A, Eguchi T, Calderwood SK. 2013. Stress proteins in aging and life span.
International Journal of Hyperthermia 29:442-447.
Müßig L, Richlitzki A, Rößler R, Eisenhardt D, Menzel R, Leboulle G. 2010. Acute
disruption of the NMDA receptor subunit NR1 in the honeybee brain selectively
impairs memory formation. Journal of Neuroscience 30:7817-7825.
Naur P, Hansen KB, Kristensen AS, Dravid SM, Pickering DS, Olsen L, Vestergaard B,
Egebjerg J, Gajhede M, Traynelis SF. 2007. Ionotropic glutamate-like receptor δ2
binds D-serine and glycine. Proceedings of the National Academy of Sciences
104:14116-14121.
Nayernia Z, Jaquet V, Krause K-H. 2014. New insights on NOX enzymes in the central
nervous system. Antioxidants & redox signaling 20:2815-2837.
Newton IG, Forbes ME, Linville MC, Pang H, Tucker EW, Riddle DR, Brunso-Bechtold
JK. 2008. Effects of aging and caloric restriction on dentate gyrus synapses and
glutamate receptor subunits. Neurobiology of aging 29:1308-1318.
Nguyen T, Hamby A, Massa SM. 2005. Clioquinol down-regulates mutant huntingtin
expression in vitro and mitigates pathology in a Huntington's disease mouse
model. Proceedings of the National Academy of Sciences of the United States of
America 102:11840-11845.
Nguyen T, Nioi P, Pickett CB. 2009. The Nrf2-antioxidant response element signaling
pathway and its activation by oxidative stress. Journal of Biological Chemistry
284:13291-13295.
Nicholson DA, Yoshida R, Berry RW, Gallagher M, Geinisman Y. 2004. Reduction in
size of perforated postsynaptic densities in hippocampal axospinous synapses and
age-related spatial learning impairments. The Journal of Neuroscience 24:76487653.
Nicoll R, Malenka R. 1999. Expression Mechanisms Underlying NMDA ReceptorDependent Long-Term Potentiation. Annals of the New York Academy of
Sciences 868:515-525.
Nicolle M, Bizon J, Gallagher M. 1996. In vitro autoradiography of ionotropic glutamate
receptors in hippocampus and striatum of aged Long–Evans rats: relationship to
spatial learning. Neuroscience 74:741-756.
Niswender CM, Conn PJ. 2010. Metabotropic glutamate receptors: physiology,
pharmacology, and disease. Annual review of pharmacology and toxicology
50:295-322.
145
Ohno S. 2013. Evolution by gene duplication: Springer Science & Business Media.
Paoletti P. 2011. Molecular basis of NMDA receptor functional diversity. European
Journal of Neuroscience 33:1351-1365.
Paoletti P, Bellone C, Zhou Q. 2013. NMDA receptor subunit diversity: impact on
receptor properties, synaptic plasticity and disease. Nature Reviews Neuroscience
14:383-400.
Paoletti P, Neyton J. 2007. NMDA receptor subunits: function and pharmacology.
Current opinion in pharmacology 7:39-47.
Paps J, Baguñà J, Riutort M. 2009. Bilaterian phylogeny: a broad sampling of 13 nuclear
genes provides a new Lophotrochozoa phylogeny and supports a paraphyletic
basal Acoelomorpha. Molecular Biology and Evolution 26:2397-2406.
Patel R, Sesti F. 2016. Oxidation of ion channels in the aging nervous system. Brain
research 1639:174-185.
Peretz B, Ringham G, Wilson R. 1982. Age-diminished motor neuronal function of
central neuron L7 in Aplysia. Journal of neurobiology 13:141-151.
Peretz B, Romanenko A, Markesbery W. 1984. Functional history of two motor neurons
and the morphometry of their neuromuscular junctions in the gill of Aplysia:
evidence for differential aging. Proceedings of the National Academy of Sciences
81:4232-4236.
Peters A. 2002. The effects of normal aging on myelin and nerve fibers: a review. Journal
of neurocytology 31:581-593.
Petralia RS, Wang YX, Zhao HM, Wenthold RJ. 1996. Ionotropic and metabotropic
glutamate receptors show unique postsynaptic, presynaptic, and glial localizations
in the dorsal cochlear nucleus. Journal of Comparative Neurology 372:356-383.
Pfaffl MW. 2001. A new mathematical model for relative quantification in real-time RT–
PCR. Nucleic acids research 29:e45-e45.
Philippe H, Lartillot N, Brinkmann H. 2005. Multigene analyses of bilaterian animals
corroborate the monophyly of Ecdysozoa, Lophotrochozoa, and Protostomia.
Molecular biology and evolution 22:1246-1253.
Pierce A, Fox B, Davis L, Visitacion N, Kitahashi T, Hirano T, Grau E. 2007. Prolactin
receptor, growth hormone receptor, and putative somatolactin receptor in
Mozambique tilapia: tissue specific expression and differential regulation by
salinity and fasting. General and comparative endocrinology 154:31-40.
Piggott BJ, Liu J, Feng Z, Wescott SA, Xu XS. 2011. The neural circuits and synaptic
mechanisms underlying motor initiation in C. elegans. Cell 147:922-933.
146
Pinsker H, Kupfermann I, Castellucci V, Kandel E. 1970. Habituation and dishabituation
of the GM-withdrawal reflex in Aplysia. Science 167:1740-1742.
Potier B, Poindessous-Jazat F, Dutar P, Billard J. 2000. NMDA receptor activation in the
aged rat hippocampus. Experimental gerontology 35:1185-1199.
Putnam NH, Butts T, Ferrier DE, Furlong RF, Hellsten U, Kawashima T, RobinsonRechavi M, Shoguchi E, Terry A, Yu J-K. 2008. The amphioxus genome and the
evolution of the chordate karyotype. Nature 453:1064-1071.
Rambaut A. 2012. FigTree version 1.4.0. Available at
http://tree.bio.ed.ac.uk/software/figtree.
Ramsköld D, Wang ET, Burge CB, Sandberg R. 2009. An abundance of ubiquitously
expressed genes revealed by tissue transcriptome sequence data. PLoS Comput
Biol 5:e1000598.
Rattan KS, Peretz B. 1981. Age-dependent behavioral changes and physiological changes
in identified neurons in Aplysia californica. Journal of neurobiology 12:469-478.
Riedel G, Platt B, Micheau J. 2003. Glutamate receptor function in learning and memory.
Behavioural brain research 140:1-47.
Roche KW, Tingley WG, Huganir RL. 1994. Glutamate receptor phosphorylation and
synaptic plasticity. Current opinion in neurobiology 4:383-388.
Rodgers KJ, Ford JL, Brunk UT. 2013. Heat shock proteins: keys to healthy ageing?
Redox Report 14:147-153.
Rohmann KN, Fergus DJ, Bass AH. 2013. Plasticity in ion channel expression underlies
variation in hearing during reproductive cycles. Current Biology 23:678-683.
Rumbaugh G, Prybylowski K, Wang JF, Vicini S. 2000. Exon 5 and spermine regulate
deactivation of NMDA receptor subtypes. Journal of neurophysiology 83:13001306.
Ryan JF, Pang K, Schnitzler CE, Nguyen A-D, Moreland RT, Simmons DK, Koch BJ,
Francis WR, Havlak P, Smith SA. 2013. The genome of the ctenophore
Mnemiopsis leidyi and its implications for cell type evolution. Science
342:1242592.
Ryu S, An H, Oh Y, Choi H, Ha M, Park S. 2008. On the role of major vault protein in
the resistance of senescent human diploid fibroblasts to apoptosis. Cell Death &
Differentiation 15:1673-1680.
Sanderson DJ, Good MA, Seeburg PH, Sprengel R, Rawlins JNP, Bannerman DM. 2008.
The role of the GluR-A (GluR1) AMPA receptor subunit in learning and memory.
Progress in brain research 169:159-178.
147
Sayers EW, Barrett T, Benson DA, Bolton E, Bryant SH, Canese K, Chetvernin V,
Church DM, DiCuccio M, Federhen S. 2011. Database resources of the national
center for biotechnology information. Nucleic acids research 39:D38-D51.
Schefe JH, Lehmann KE, Buschmann IR, Unger T, Funke-Kaiser H. 2006. Quantitative
real-time RT-PCR data analysis: current concepts and the novel “gene
expression’s CT difference” formula. Journal of molecular medicine 84:901-910.
Schenk D, Barbour R, Dunn W, Gordon G, Grajeda H, Guido T, Hu K, Huang J,
Johnson-Wood K, Khan K. 1999. Immunization with amyloid-β attenuates
Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400:173-177.
Schneggenburger R. 1996. Simultaneous measurement of Ca2+ influx and reversal
potentials in recombinant N-methyl-D-aspartate receptor channels. Biophysical
journal 70:2165-2174.
Schwartz JH, Castellucci VF, Kandel ER. 1971. Functioning of identified neurons and
synapses in abdominal ganglion of Aplysia in absence of protein synthesis.
Journal of Neurophysiology 34:939-953.
Shankar GM, Li S, Mehta TH, Garcia-Munoz A, Shepardson NE, Smith I, Brett FM,
Farrell MA, Rowan MJ, Lemere CA. 2008. Amyloid-β protein dimers isolated
directly from Alzheimer's brains impair synaptic plasticity and memory. Nature
medicine 14:837-842.
Shi L, Adams MM, Linville MC, Newton IG, Forbes ME, Long AB, Riddle DR, BrunsoBechtold JK. 2007. Caloric restriction eliminates the aging-related decline in
NMDA and AMPA receptor subunits in the rat hippocampus and induces
homeostasis. Experimental neurology 206:70-79.
Shigenaga MK, Hagen TM, Ames BN. 1994. Oxidative damage and mitochondrial decay
in aging. Proceedings of the National Academy of Sciences 91:10771-10778.
Simpson JH. 2009. Mapping and manipulating neural circuits in the fly brain. Advances
in genetics 65:79-143.
Sisneros JA, Bass AH. 2003. Seasonal plasticity of peripheral auditory frequency
sensitivity. The Journal of Neuroscience 23:1049-1058.
Sobolevsky AI, Rosconi MP, Gouaux E. 2009. X-ray structure, symmetry and
mechanism of an AMPA-subtype glutamate receptor. Nature 462:745-756.
Song I, Huganir RL. 2002. Regulation of AMPA receptors during synaptic plasticity.
Trends in neurosciences 25:578-588.
Stadtman ER. 2001. Protein oxidation in aging and age-related diseases. Annals of the
New York Academy of Sciences 928:22-38.
148
Stern P, Behe P, Schoepfer R, Colquhoun D. 1992. Single-channel conductances of
NMDA receptors expressed from cloned cDNAs: comparison with native
receptors. Proceedings of the Royal Society of London B: Biological Sciences
250:271-277.
Stommes D, Fieber LA, Beno C, Gerdes R, Capo TR. 2005. Temperature effects on
growth, maturation, and lifespan of the california sea hare (Aplysia californica).
Journal of the American Association for Laboratory Animal Science 44:31-35.
Strimmer K. 2008. fdrtool: a versatile R package for estimating local and tail area-based
false discovery rates. Bioinformatics 24:1461-1462.
Tamaru M, Yoneda Y, Ogita K, Shimizu J, Nagata Y. 1991. Age-related decreases of the
N-methyl-D-aspartate receptor complex in the rat cerebral cortex and
hippocampus. Brain research 542:83-90.
Team RC. 2014. R: A language and environment for statistical computing. R Foundation
for Statistical Computing, Vienna, Austria. 2013. In.
Teng H, Cai W, Zhou L, Zhang J, Liu Q, Wang Y, Dai W, Zhao M, Sun Z. 2010.
Evolutionary mode and functional divergence of vertebrate NMDA receptor
subunit 2 genes. PLoS One 5:e13342.
Tichelaar W, Safferling M, Keinänen K, Stark H, Madden DR. 2004. The threedimensional structure of an ionotropic glutamate receptor reveals a dimer-ofdimers assembly. Journal of molecular biology 344:435-442.
Toescu E, Myronova N, Verkhratsky A. 2000. Age-related structural and functional
changes of brain mitochondria. Cell calcium 28:329-338.
Traynelis SF, Wollmuth LP, McBain CJ, Menniti FS, Vance KM, Ogden KK, Hansen
KB, Yuan H, Myers SJ, Dingledine R. 2010a. Glutamate receptor ion channels:
structure, regulation, and function. Pharmacological reviews 62:405-496.
Traynelis SF, Wollmuth LP, McBain CJ, Menniti FS, Vance KM, Ogden KK, Hansen
KB, Yuan H, Myers SJ, Dingledine R. 2010b. Glutamate receptor ion channels:
structure, regulation, and function. Pharmacol Rev 62:405-496.
Trudeau L, Castellucci VF. 1995. Postsynaptic modifications in long-term facilitation in
Aplysia: upregulation of excitatory amino acid receptors. The Journal of
Neuroscience 15:1275-1284.
Trudeau L-E, Castellucci VF. 1993. Excitatory amino acid neurotransmission at sensorymotor and interneuronal synapses of Aplysia californica. Journal of
Neurophysiology 70:1221-1230.
Tsai G, Coyle JT. 2002. Glutamatergic mechanisms in schizophrenia. Annual review of
pharmacology and toxicology 42:165-179.
149
Uttara B, Singh AV, Zamboni P, Mahajan R. 2009. Oxidative stress and
neurodegenerative diseases: a review of upstream and downstream antioxidant
therapeutic options. Current neuropharmacology 7:65-74.
Villareal G, Li Q, Cai D, Glanzman DL. 2007. The role of rapid, local, postsynaptic
protein synthesis in learning-related synaptic facilitation in Aplysia. Current
Biology 17:2073-2080.
Vogel C, Marcotte EM. 2012. Insights into the regulation of protein abundance from
proteomic and transcriptomic analyses. Nature Reviews Genetics 13:227-232.
Wallace M, Frankfurt M, Arellanos A, Inagaki T, Luine V. 2007. Impaired recognition
memory and decreased prefrontal cortex spine density in aged female rats. Annals
of the New York Academy of Sciences 1097:54-57.
Walters ET, Byrne J, Carew T, Kandel E. 1983a. Mechanoafferent neurons innervating
tail of Aplysia. I. Response properties and synaptic connections. Journal of
Neurophysiology 50:1522-1542.
Walters ET, Byrne J, Carew T, Kandel E. 1983b. Mechanoafferent neurons innervating
tail of Aplysia. II. Modulation by sensitizing stimulation. Journal of
Neurophysiology 50:1543-1559.
Wang W, Simovic DD, Di M, Fieber L, Rein KS. 2013. Synthesis, receptor binding and
activity of iso and azakainoids. Bioorganic & medicinal chemistry letters
23:1949-1952.
Wardas J, Pietraszek M, Schulze G, Ossowska K, Wolfarth S. 1996. Age-related changes
in glutamate receptors: an autoradiographic analysis. Polish journal of
pharmacology 49:401-410.
Wenk GL, Barnes CA. 2000. Regional changes in the hippocampal density of AMPA and
NMDA receptors across the lifespan of the rat. Brain research 885:1-5.
Wollmuth LP, Sobolevsky AI. 2004. Structure and gating of the glutamate receptor ion
channel. Trends Neurosci 27:321-328.
Xia S, Miyashita T, Fu T-F, Lin W-Y, Wu C-L, Pyzocha L, Lin I-R, Saitoe M, Tully T,
Chiang A-S. 2005. NMDA receptors mediate olfactory learning and memory in
Drosophila. Current Biology 15:603-615.
Yankner BA, Lu T, Loerch P. 2008. The aging brain. Annu. Rev. pathmechdis. Mech.
Dis. 3:41-66.
Zamanillo D, Sprengel R, Hvalby Ø, Jensen V, Burnashev N, Rozov A, Kaiser KM,
Köster HJ, Borchardt T, Worley P. 1999. Importance of AMPA receptors for
hippocampal synaptic plasticity but not for spatial learning. Science 284:18051811.
150
Zeddies DG, Fay RR, Alderks PW, Shaub KS, Sisneros JA. 2010. Sound source
localization by the plainfin midshipman fish, Porichthys notatus. The Journal of
the Acoustical Society of America 127:3104-3113.
Zhao S, Fernald RD. 2005. Comprehensive algorithm for quantitative real-time
polymerase chain reaction. Journal of computational biology 12:1047-1064.
Zhao Y, Hegde AN, Martin KC. 2003. The ubiquitin proteasome system functions as an
inhibitory constraint on synaptic strengthening. Current Biology 13:887-898.
APPENDIX A
CHAPTER 2 SUPPLEMENTARY FIGURES
1
0.278Grin2A_Human
Grin2A_Rat
Grin2A_Mouse
Grin2A_Zebrafish
Grin2B_Zebrafish
0.972
1 Grin2B_Rat
0.308
Grin2B_Human
Grin2B_Mouse
Grin2C_Zebrafish
0.844
Grin2C_Mouse
0.998
1
Grin2C_Rat
0.564
Grin2C_Human
0.934
0.97
1
1
0.966
1
0.886
Grin2D_Zebrafish
Grin2D_Mouse
0.27Grin2D_Human
0.948
1
0.324
0.106
0.158
1
1
1
0.928
0.894
0.996
0.528
0.382
0.982
0.36
1
0.948
0.536
0.824
Grin2D_Rat
Grin2A_Ciona
Grin2_Lancelet
Grin2_Aplysia
Grin2_Celegans
Grin2_Priapulus
Grin2_Lingula
Grin2_Limulus
Grin2_Daphnia
Grin2_Drosophila
1
1
Grin1_Tribolium
Grin2_Tribolium
Grin1_Celegans
Grin1_Lingula
Grin1_Drosophila
Grin1_Daphnia
Grin1_Limulus
Grin1_Priapulus
Grin1_Aplysia
Grin1_Octopus
Grin1_Zebrafish
0.59 Grin1_Human
0.818Grin1_Mouse
Grin1_Rat
Grin3_Daphnia
Grin3_Limulus
Grin3_Octopus
0.908
Grin3_Priapulus
1
1
1
Grin3A_Lancelet
Grin3A_Zebrafish
Grin3A_Human
0.998 Grin3A_Mouse
Grin3A_Rat
0.2
Fig. S2.1. NMDAR subtype only tree. Phylogeny using sequences identified in the full
phylogeny as NMDAR. Protostome sequences show monophyletic relationships with
chordates for each subunit as seen in the full phylogeny. After alignment trimming there
were 467 positions used for analysis. Numbers indicated bootstrap support and the scale
bar represent 0.2 substitutions per site.
151
152
GluR3_Aplysia
0.86
GluR4_Aplysia
0.52
GluR_Octopus
GluR3_Priapulus
0.95
0.96
GluR4_Priapulus
0.62
GluR1_Limulus
1
GluR1A_Drosophila
1
GluRIB_Drosophila
0.16
GluR5_Aplysia
1
GluR8_Aplysia
0.72
GluR4_Lingula
0.44
GluR1_Aplysia
1
0.18
GluR2_Aplysia
GluR2_Lingula
0.12
GluR1_Celegans
1
GluR2_Celegans
0.88 GRIA1_Mouse
1
GRIA1_Rat
1
GRIA1_Human
GRIA1_Zebrafish
0.41
GRIA2_Zebrafish
1
1
GRIA2_Human
0.71GRIA2_Mouse
GRIA2_Rat
1
1 GRIA3_Human
1
1
0.96
GRIA3_Rat
GRIA3_Mouse
GRIA3_Zebrafish
1
GRIA4_Zebrafish
1
0.91
GRIA4_Mouse
0.02GRIA4_Human
GRIA4_Rat
GluR1_Ciona
0.2
Fig. S2.2. AMPAR subtype only tree. Phylogeny of sequences identified as AMPAR
genes in the full phylogeny. The tree indicates a single AMPAR gene in the common
bilaterian ancestor as seen in the full phylogeny. After alignment trimming there were
665 positions used for analysis. Numbers indicated bootstrap support and the scale bar
represent 0.2 substitutions per site.
153
KA2_Aplysia
0.96
GRIK2_Limulus
1
GRIK1_Tribolium
0.94
GRIK2_Lingula
GluR7_Aplysia
0.53
GluR2D_Drosophila
0.83
1
GluRIIE_Drosophila
0.84
GRIK3_Limulus
0.72
GRIK1_Daphnia
0.57
GluR1_Tribolium
GRIK2_Human
0.42
GRIK2_Rat
1
GRIK2_Mouse
0.99
GRIK2_Zebrafish
0.62
GRIK3_Zebrafish
1
GRIK3_Human
1
GRIK3_Mouse
0.84
GRIK3_Rat
1
GRIK1_Zebrafish
1
GRIK1_Human
1
0.94
GRIK1_Mouse
0.99
GRIK1_Rat
GRIK2_Lancelet
GRIK4_Mouse
0.92
GRIK4_Rat
0.63
1
1
GRIK4_Human
GRIK4_Zebrafish
1
GRIK5_Zebrafish
0.96
GRIK5_Human
1
GRIK5_Mouse
0.89
GRIK5_Rat
0.2
Fig. S2.3. Kainate receptor subtype only tree. Phylogeny of sequences identified as
kainate receptor genes in the full tree. There were 687 positions in the alignment after
trimming. Scale bar represents 0.2 amino acid substitutions per site.
154
GluR1_Aplysia
0.994
GluR2_Aplysia
0.873
GluR5_Aplysia
0.995
GluR8_Aplysia
0.9
GluR3_Aplysia
0.788
0.99
GluR4_Aplysia
GluR7_Aplysia
0.975
KA2_Aplysia
GluR6_Aplysia
0.941
KA1_Aplysia
Grin2_Aplysia
1
Grin1-1_Aplysia
1
Grin1-2_Aplysia
0.2
Fig. S2.4. Aplysia iGluR protein tree. Subunits within a clade can form functional
receptors in vertebrates, and clades formed by both AMPAR and NMDAR-type subunits
of Aplysia suggest subunits within these clades may form complete receptors. Kainate
receptor subunits do not form a clade. GluR7 and KA2 are the more likely kainate
receptor subunit candidates based on their positions full tree. Numbers indicate bootstrap
support and the scale bar represents 0.2 substitutions per site.
155
3.5
TMD1
TMD2
TMD3
3
Hydrophobicity score
2.5
2
Aplysia GluR1
Human GRIA1
1.5
1
0.5
0
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37
Fig. S2.5. Representative AMPA subunit TMD Hydrophobicity. Aplysia GluR1 is
monophyletic with vertebrate AMPAR subunits in the phylogeny, and shows similar
hydrophobicity in its TMD1 and TMD3 to Human GRIA1. This suggests similarities in
the protein structure of the ion channel of these two subunits.
156
3.5
TMD1
TMD2
TMD3
Hydrophobicity Score
3
2.5
2
Aplysia GluR7
Human GRIK1
1.5
1
0.5
0
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
Fig. S2.6. Representative kainate subunit TMD Hydrophobicity. Aplysia GluR7 is
monophyletic with vertebrate kainate receptor subunits in the phylogeny, and shows
similar hydrophobicity in its TMDs to Human GRIK1. This suggests similarities in the
protein structure of the ion channel of these two subunits.
APPENDIX B
Chapter 4 Supplementary Data Files
Description:
The accompanying Excel spreadsheets show the full datasets for complete gene ontology
terms (Supplementary File 1) and differential expression results (Supplementary File 2)
from RNASeq analysis.
Filenames:
Chapter_4_Supp_File_1.xlsx
Chapter_4_Supp_File_2.xlsx
157