Investigating the influence of LH-projecting BLA neurons
upon motivated behavioral responding and appetitive learning
ARCHES
MASSACHUSETTS INSTITUTE
OF TECHNOLOGY
by
JUL 2 2 2015
Christine Eckhardt
A.B., Neurobiology
Harvard College, 2009
LIBRARIES
Submitted to the MIT Department of Brain and Cognitive Sciences in Partial
Fulfillment of the Requirements for the Degree
of
Master of Science in Brain and Cognitive Sciences
at the
Massachusetts Institute of Technology
June 2015
c 2015 Massachusetts Institute of Technology. All rights reserved.
Signature of Author
Signature redacted__
MIT
Department of Brain and Cognitive Sciences
May 4, 2015
Certified by
Signature redacted,
Kay Tye
Asst~raPr essor, 4Power Insitute of Learning and Memory, MIT
\/
Thesis Supervisor
Accepted by_
Signature red acted~-
Professor, Department of Brain and Cognitive Sciences
Director of Graduate Education for Brain and Cognitive Sciences
Investigating the role of LH-projecting BLA neurons
in motivated behavioral responding and appetitive learning
by
Christine Eckhardt
Submitted to the Department of Brain & Cognitive Sciences on May 4, 2015
In Partial Fulfillment of the Requirement for the Degree of
Master of Science in Brain and Cognitive Sciences
Abstract
To optimize survival, organisms must be able to learn contingencies between external
stimuli and rewards and appropriately respond to these associations. Deficits in reward-related
learning or reward-seeking are thought to occur in a host of psychopathologies, including
depression (Drevets, 2001), eating disorders (Wagner et al., 2007), and substance abuse (Wrase
et al., 2007), such that improved understanding of reward processing could potentially aid in the
development of therapies. Two neural regions, the basolateral amygdala (BLA) and lateral
hypothalamus (LH), are both implicated in reward processing (Adamantidis et al., 2007; Anand
and Brobeck, 1951; Brobeck, 1946; Gutierrez et al., 2011; Hoebel and Teitelbaum, 1962;
Kempadoo et al., 2013; Margules and Olds, 1962; Muramoto et al., 1993; Sakurai, 2007;
Schoenbaum et al., 1998; Tye and Janak, 2007; Tye et al., 2008, 2010), but the role of the BLA's
projection to LH in appetitive conditioning and reward-seeking remains unclear. Through the use
of optogenetic techniques in mice, I have investigated the influence of LH-projecting BLA
neurons upon motivated behavioral responding, which has indicated that the projection may
support intracranial self-stimulation (ICSS). Further experiments with in vivo extracellular
electrophysiological recordings from LH-projecting BLA neurons may also shed light on the
encoding properties of these neurons during appetitive learning.
Thesis Supervisor: Kay Tye
Title: Assistant Professor Neuroscience
Acknowledgements
Thank you to all the members of the Tye Lab for their advice and assistance over the
years, and Dr. Kay Tye for her support of this research. Thank you to Melodi Anahtar for being
the best undergraduate a graduate student could hope to train, and, in particular, thank you to
Anna Beyeler for her unwavering enthusiasm and willingness to help. Thank you as well to Dr.
Matt Wilson for his guidance in my graduate school trajectory, and the BCS department for their
support. I also greatly appreciate the advice of Dr. Frosch, Dr. Feany, and Dr. Mitchell at
Harvard Medical School over the past six years, and their continuing encouragement in pursuing
a career as academic-physician scientist.
Lastly, above all, thank you to my husband, Priya, Morgan, and my family, for giving me
a strong backbone to support my professional life. Thank you for always listening.
3
Table of Contents
Title Page
Abstract
Acknowledgements
Table of Contents
List of Figures
1. Introduction
1.1
1.2
1.3
1.4
1.5
1.6
Implication of BLA in appetitive learning
Role of the LH in reward-related processes
Identification of the BLA to LH projection
LH-projecting BLA neurons respond to appetitive conditioned stimuli
Contribution of LH-projecting BLA neurons to BLA functions remains unknown
Investigating the role of LH-projecting BLA neurons in positive reinforcement
2. Methods
2.1
2.2
2.3
2.4
2.5
2.6
3.
Animals and Stereotaxic Surgery
In vivo optogenetic manipulation
Behavioral assays
Visualizing viral transduction and fiber optic placement
Statistical Analysis
In vivo extracellular electrophysiology
Results
Optical stimulation of BLA fibers in LH supports ICSS
ICSS most likely arises from reward-related processes
Novel object exploration & feeding behavior in a subset of ChR2 animals
Absence of optical self-stimulation with putative transduction of LH-projecting
BLA neurons
3.5 Feasibility of in vivo electrophysiological recordings during operant appetitive
conditioning
3.1
3.2
3.3
3.4
4. Discussion
1
2
3
4
5
6
6
9
10
11
11
12
13
13
17
18
23
24
24
27
27
27
30
30
35
40
4.1 Projection from BLA to LH may support motivated behavioral responding
4.2 Analysis of excluded animals indicates possible opposing effect of CEA to LH
projection
40
41
4.3 Activation of BLA fibers in LH do not nonspecifically trigger feeding
42
4.4 LH anatomy necessitates more specific optogenetic methods
4.5 Preliminary results with cre-DIO method
4.6 Future directions for investigating the encoding properties of LH-projecting
BLA neurons
4.7 Possible firing patterns among LH-projecting BLA neurons
43
44
45
46
5.
Conclusion
49
6.
References
50
4
List of Figures
Stereotaxic surgery for optogenetic manipulations and recordings.
Behavioral assays.
Confocal imaging reveals viral transduction of BLA and fiber placement over LH.
Optical stimulation of BLA fibers supports ICSS.
ICSS in ChR2 animals does not arise from a locomotor effect.
Anxiety-related behaviors do not appear to contribute to ICSS in ChR2 animals.
Preliminary results of novel object exploration and feeding with optical stimulation.
Absence of object preference among naive wildtype animals.
No significant effect of optical stimulation of putative LH-projecting BLA neurons.
Successful transduction of LH-projecting BLA neurons with ChR2-eYFP and
in vivo extracellular recordings.
11. Behavioral learning curve displays animal's acquisition of the task.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
14
20
28
29
31
32
33
34
36
37
39
S
Introduction
In order to survive, an organism must be able to learn and appropriately respond to
environmental contingencies between stimuli and rewards. Perturbations to reward processing
are thought to contribute to numerous psychopathologies, including depression (Drevets, 2001),
eating disorders (Wagner et al., 2007), and addiction (Wrase et al., 2007), such that enhanced
understanding of the biological mechanisms of reward processing could have tremendous
therapeutic benefit. The basolateral amygdala (BLA) has been shown to play a critical role in the
association of environmental cues and appetitive outcomes (Muramoto et al., 1993; Schoenbaum
et al., 1998; Tye and Janak, 2007; Tye et al., 2008, 2010) as well as the ability of these cues to
gain motivating properties (Tye et al., 2010). Despite these findings, the functions of the BLA's
efferent connection to the lateral hypothalamus (LH) in reward processing are not entirely
known, as pre-existing techniques such as electrophysiology and electrical stimulation did not
permit targeting of specific projections. Through the use of optogenetic techniques, I have
investigated the contribution of the projection from the BLA to LH to motivated behavioral
responding, in order to shed light on this connection's possible role in appetitive learning and
reward-seeking behavior.
Implication of BLA in Appetitive Learning
The BLA is a cortical-like structure of predominantly glutamatergic projection neurons
that includes three nuclei: the lateral amygdala (LA), basolateral amygdala (BL), and basomedial
amygdala (BMA) (Pape and Pare, 2010). Although substantial work has focused on the BLA and
fear conditioning (LeDoux et al., 1990; Maren and Quirk, 2004; Pape and Pare, 2010; Quirk et
al., 1995, 1997; Wilensky et al., 1999), the region has also been implicated in the association of
cues, or conditioned stimuli (CS), and appetitive outcomes, or unconditioned stimuli (US).
Electrophysiological recordings in rats and non-human primates (NHP) have revealed a subset of
6
BLA neurons that respond to reward-predictive cues (Muramoto et al., 1993; Sanghera et al.,
1979; Tye and Janak, 2007; Uwano et al., 1995), and a subpopulation of BLA neurons is thought
to represent the value of reward-predictive cues (Baxter and Murray, 2002; Belova et al., 2008;
Paton et al., 2006; Schoenbaum et al., 2007). In NHPs, these neurons show a graded response
depending on the magnitude of the reward, as well as activity that tracks reward-predictive cues
following any changes in cue-outcome pairings (Belova et al., 2008; Paton et al., 2006). Another
subpopulation exhibits a similar pattern of activity, but in response to aversive outcomes,
indicating both positive and negative coding cells in the BLA.
In addition to representing value, the BLA may also endow a reward-predictive cue with
reinforcing properties (Baxter and Murray, 2002). In the extinction phase of an operant
appetitive conditioning task, rats will continue to nosepoke to trigger a cue, even after they
discontinue checking for an appetitive outcome (Tye and Janak, 2007). This persistent
nosepoking, long after the cue ceases to predict an outcome, indicates that the cue has acquired
reinforcing properties. The firing of a subset of BLA neurons correlates with this behavior,
indicating the BLA's involvement in supporting reinforcement.
Consistent with this implication, BLA lesions have also produced deficits in 2nd-order
conditioning and reinforcer devaluation (Hatfield et al., 1996; Milkovi et al., 1997). In 2nd-order
conditioning, a CS1 previously paired with a US is able to act as a reinforcer, causing an animal
to respond to a CS2 paired with CS1, even in the absence of the original US. BLA lesions impair
2ndorder, but not 1-order, conditioning, indicating that the region contributes to the ability of
the CS to act as a reinforcer. In reinforcer devaluation, an animal learns that a CS predicts a US
(food), and the US is later devalued through pairing of the food with malaise, in the absence of
the CS. BLA lesions also impair reinforcer devaluation, suggesting that the BLA is involved in
updating the value of the CS. Thus, the BLA is involved in motivated behavioral responding, via
its possible endowment of cues with reinforcing properties and signaling of incentive value of
the cue.
As found with fear conditioning (Maren, 2005; McKernan and Shinnick-Gallagher, 1997;
Rogan et al., 1997; Rumpel et al., 2005), cellular and synaptic changes consistent with long term
potentiation occur alongside cue-reward learning in the BLA. After an appetitive operant (Tye et
al., 2008) or Pavlovian auditory conditioning task (Namburi et al., 2015), a subset of BLA
neurons exhibited potentiation of putative thalamic synapses, reflected by an increase in the
AMPA/NMDA
ratio, which is thought to indicate glutamatergic synaptic strength. This
potentiation is considered NMDA-dependent, given that infusion of an NMDA-receptor
antagonist prior to training diminished reward-learning performance (Tye et al., 2008). Blockade
of NMDA-receptors also reduced the amplitude of mini excitatory postsynaptic currents
(mEPSCs) in ex vivo slice recordings following training, as compared to animals receiving
vehicle alone, which is consistent with decreased AMPA receptor number or function. Other
studies have similarly reported impaired acquisition of Pavlovian appetive tasks with infusion of
an NMDA-receptor antagonist into the BLA (Burns et al., 1994). However, lesion studies of the
BLA have not reported any deficits in Is'-order Pavlovian appetitive conditioning (Hatfield et al.,
1996; Holland et al., 2001), although adaptation or redundant circuitry may account for this
discrepancy.
Given the overlap in inputs that show potentiation in response to appetitive and aversive
conditioning, the valence of a given BLA neuron may arise from its projection target (Namburi
et al., 2015). Indeed, BLA neurons that project to the nucleus accumbens (NAc), a region
implicated in reward-related processes (Caine et al., 1995; Hurd et al., 1989; Pettit and Justice
Jr., 1989), show bidirectional modulation of potentiation depending on the type of conditioning,
with an increase in the AMPA/NMDA ratio after appetitive learning, and a decrease after
S8
aversive learning (Namburi et al., 2015). Inverse changes in AMPA/NMDA ratio are observed in
BLA neurons projecting to the central medial nucleus of amygdala, which is heavily implicated
in aversive conditioning (Ciocchi et al., 2010; Haubensak et al., 2010; Jimenez and Maren,
2009). Thus, investigation of the BLA according to projection target can yield important insight
into the region's contribution to associative learning, yet the role of the BLA's connection to LH,
another region implicated in reward, remains unclear.
Role of the LH in reward-relatedprocesses
The LH is a large and remarkably heterogenous region that has been implicated in
reward, motivation, and consummatory behaviors (Adamantidis et al., 2007; Anand and
Brobeck, 1951; Brobeck, 1946; Gutierrez et al., 2011; Harris et al., 2005; Hoebel and
Teitelbaum, 1962; Kempadoo et al., 2013; Margules and Olds, 1962; Sakurai, 2007). Early
studies of LH demonstrated intracranial self stimulation (ICSS) with electrical stimulation of LH
(Olds and Milner, 1954), as well as grooming and sexual behaviors (Singh et al., 1996), and
extracellular recordings revealed neurons responsive to reward-associated cues (Mora et al.,
1976; Schwartzbaum, 1988). Further work has shown populations of LH neurons that encode
reward predictive cues, reward retrieval, and unexpected omission of reward (Nieh et al., 2015).
The projection of the LH to ventral tegmental area (VTA), a crucial region in reward
learning (Schultz, 2006; Schultz et al., 1997), has been found to contribute to reward-related
behaviors. Electrical (Bielajew and Shizgal, 1986) and optogenetic (Kempadoo et al., 2013)
activation of this projection supports ICSS, and these neurons appear to encode the learned
action of pursuing an appetitive outcome, even if the outcome is unavailable (Nieh et al., 2015).
LH neurons responsive to reward-predictive cues and unexpected omission of reward appear to
be downstream of VTA neurons that receive input from LH, and they may receive information
from the VTA either directly via a reciprocal connection or indirectly, via the nucleus
accumbens, hippocampus, ventral pallidum, and amygdala (Barone et al., 1981; Beckstead et al.,
1993; Simon et al., 1979). Thus, the BLA could potentially convey information from the VTA
and/or thalamic inputs among others, necessitating further study of the BLA to LH projection.
Identification of the BLA to LHprojection
The connection between BLA and LH was identified through histological and viral
techniques. The projection was first labelled via an anterograde tracer, Phaseolus vulgaris
leucoagglutinin (PHAL), in rats, which showed an efferent connection to LH from posterior
BLA and anterior BMA, as well as adjacent nuclei, including CEA, anterior amygdaloid and
cortical amygdaloid nuclei. Through the use of biotinylated dextranamine and immunostaining
for melanin concentrating (MCH), BMA/CoA neurons were found to project directly onto MCH+
neurons in the LH (Niu et al., 2012). Researchers also demonstrated that the BLA, BMA, and
BMP project either directly or indirectly onto orexin neurons in LH by expressing a retrograde
tracer, a tetanus toxin fused to GFP, in orexin neurons. However, the tracer's ability to jump
more than one synapse precludes definitive determination of monosynaptic projection (Sakurai et
al., 2005). Another study employing a pseudorabies virus with Cre-dependence also indicated
that BLA neurons may project to leptin & neuropeptide-Y containing neurons (DeFalco et al.,
2001), although, again, the virus employed could travel more than one synapse retrogradely.
While more precise methods are needed to determine the monosynaptic connections of LHprojecting BLA neurons, these studies suggest the potential influence of the BLA upon
neuropeptides crucial to energy homeostasis and arousal (Berthoud and Mutnzberg, 2011). Given
that appetitive learning often pertains to unconditioned stimuli that affect energy, this projection
from BLA to LH, as well as the reciprocal connection from LH to BLA, may provide an
important link in the valuing of stimuli in relation to homeostasis.
I0
LH-projecting BLA neurons respond to appetitive conditionedstimuli
Although, as discussed, extensive work has implicated the BLA and LH in reward-related
processes (Adamantidis et al., 2007; Anand and Brobeck, 1951; Brobeck, 1946; Gutierrez et al.,
2011; Hoebel and Teitelbaum, 1962; Kempadoo et al., 2013; Margules and Olds, 1962;
Muramoto et al., 1993; Sakurai, 2007; Schoenbaum et al., 1998; Tye and Janak, 2007; Tye et al.,
2008, 2010), the efferent connection between BLA and LH has only been investigated through
lesion and histological studies (Holland et al., 2002; Petrovich and Gallagher, 2003; Petrovich et
al., 2002, 2005). The projection is thought to contribute to CS-potentiated feeding (Holland et al.,
2002; Petrovich et al., 2002, 2005), a phenomenon in which a CS will trigger eating in sated
animals who previously ignored food, provided the animal learned the association of CS and
food in a previous food-restricted state (Zamble, 1973). The assay may capture processes that
contribute to overeating and obesity in humans (Petrovich and Gallagher, 2003; Rodin, 1981).
Bilateral lesions of the BLA, but not the central amygdala (CEA), abolishes CS-potentiated
feeding, without affecting acquisition of the CS-US pairing or baseline food consumption
(Holland et al., 2002). Contralateral, but not ipsilateral, lesions of the BLA and LH similarly
impair CS-potentiated feeding (Petrovich et al., 2002). Through the use of a retrograde tracer and
staining for immediate early genes, researchers also reported activation of BMA/BLA neurons
that project to LH in response to the CS t in CS-potentiated feeding, far more than another cue
that was paired to no outcome (CS-) (Petrovich et al., 2005). This indicates that LH-projecting
BLA neurons may evoke the value of the CS (Petrovich and Gallagher, 2003), potentially as it
relates to energy homeostasis.
Contributionof LH-projectingBLA neurons to BLA functions remains unknown
Beyond these studies of CS-potentiated feeding, little is known about the projection's
contribution to other functions attributed to the BLA. BLA and LH contralateral lesions do not
i
impair 2nd-order conditioning, but more temporally precise manipulations could reveal an effect,
similar to that found with the projection from BLA to NAc (Setlow et al., 2002). Different BLA
projections have been found to promote or diminish anxiety-related behaviors, with BLA to
ventral hippocampus exhibiting an anxiogenic influence (Felix-Ortiz et al., 2013) and BLA to
CeA an anxiolytic effect (Tye et al., 2011). However, any effect of LH-projecting BLA neurons
upon anxiety-related behaviors remains unclear. The projection from BLA to NAc supports ICSS
(Stuber et al., 2011), yet the effect of LH-projecting BLA neurons upon this process also has not
been studied. Given the role of both BLA and LH in reward-related processes, investigation of
this projection in the context of reward-seeking and appetitive learning is warranted, as well as
control studies of anxiety-related behaviors.
Investigatingthe role ofLH-projectingBLA neurons in positive reinforcement
Through optogenetic techniques, I have investigated the projection from BLA to LH in a
temporally and spatially specific manner, in order to identify the influence of these neurons upon
motivated behavioral responding, as well as support future studies of this projection's role in
appetitive associative learning. I have found preliminary evidence that BLA to LH supports
ICSS, without pronounced effects upon locomotion or anxiety-related behaviors. Given that
activation of LH neurons supports ICSS, it is likely that more precise experiments will verify
these results, but further controls are required to substantiate my findings. I have also
demonstrated that identification of LH-projecting BLA neurons in in vivo extracellular
recordings is possible, and that such recordings could be conducted during acquisition of an
appetitive conditioning task in order to study the encoding properties of these neurons.
12
Methods
All procedures were performed according to the guidelines from the NIH and with
approval of the MIT IACUC and DCM.
Animals and Stereotaxic Surgery
For all stereotaxic surgeries, adult wildtype male C57BL/6 mice (8.3 h 1.5 weeks;
Jackson Laboratory, Bar Harbor, ME) were used. All surgeries were performed under aseptic
conditions with a digital small animal stereotaxic instrument (David Kopf Instruments, Tujunga,
CA), and all mice were anaesthetized with isoflurane (5 % for induction, 1.5-2.0 % after) during
surgery. A heating pad was used to maintain body temperature throughout surgery, and
afterwards, animals were subjected to a heating lamp until fully recovered. Following recovery,
animals were housed in a 12 hour reverse light-dark cubicle (darkness from 9 am to 9 pm) to
promote wakefulness during behavioral assays.
Viral transductionof BLA neurons and opticfiber implantationover LH
In order to manipulate the activity of LH-projecting BLA neurons with optical
stimulation, wildtype mice underwent stereotaxic surgery with viral infusion in the BLA and
optic fiber implantation over the LH (Fig. IA). I infused the BLA unilaterally with .425 tl of
adeno-associated virus (AAV 5) carrying the opsin fused to a fluorophore, hChR2(H134R)-eYFP,
under a Ca
calmodulin-dependent protein kinase IIa (CaMKIla) promotor at stereotaxic
coordinates from bregma: -1.5 mm anteroposterior (AP), +3.35 mm mediolateral (ML) and -4.89
mm dorsoventral (DV). This virus has been shown to produce expression of ChR2 in
glutamatergic BLA neurons (Tye et al., 2011). In a second group of animals, I infused a
fluorophore control virus (AAV5 -CaMKIla -eYFP), and surgery upon ChR2 and eYFP animals
was interleaved throughout the day. For each group of animals, I used a different NanoFil
syringe (10 pL; WPI, Sarasota, FL) with a beveled 33 gauge needle and infused the virus at a
A
AAV-CaMKIlctChR2-eYFP
Behavior
4-5 weeks
LLH
BLBA
BLA
CAV2-Cre
AAV5-DIOChR2-eYFP
Behavior
6 weeks
LH
LH
BLA
B
CAV2-Cre
AAV 5-DIOChR2-eYFP
Recordings during
behavior & "phototagging"
6 weeks
LH
LA
BLA
Figure 1. Stereotaxic surgery for optogenetic manipulations and recordings.
A, BLA neurons are transduced with ChR2-eYFP, and BLA terminals in LH are activated during
behavior via stimulation with blue light (473 nm). B, LH-projecting BLA neurons are transduced
with ChR2-eYFP through infusion of a retrograde cre-recombinase virus in LH and a cre-dependent
virus carrying ChR2-eYFP in the BLA. An optic fiber is implanted over the BLA to activate LHprojecting BLA neurons during behavior. C, LH-projecting BLA neurons are transduced with
ChR2-eYFP using the same method as B, followed by optrode implantation in the BLA for
extracellular recordings during behavior.
14
rate of .1 IL /min using a microsyringe pump (UMP3; WPI, Sarasota, FL) and controller
(Micro4; WPI, Sarasota, FL). After infusion, the syringe was raised 50 Pm and left for twenty
minutes to allow diffusion of virus and guard against leakage of virus along the needle tract.
After viral infusion, a 300 jim optic fiber (0.37 numerical aperture, NA) glued to a
2.5 mn stainless steel ferrule (ThorLabs) was implanted over the LH (-.5 mm AP, +1.05 mm
ML, AP -. 5, 4.85 mm DV). The pre-implantation efficiency of all fibers was greater than 80%.
Adhesive cement (C&B metabond; Parkell, Edgewood, NY) was applied to the skull and fiber,
followed by cranioplastic cement (Dental cement; Stoelting, Wood Dale, IL) to anchor the
implant. Twenty minutes following cement application, the incision was closed with nylon
sutures. Based on optogenetic studies with the same virus in our lab, animals were given 4-5
weeks post-surgery with ad libitum food and water to insure sufficient opsin in BLA terminals
during behavioral assays. Surgeries for two of these animals, one of which was excluded due to
viral leak, was performed by a labmate, Chris Leppla.
Specific viral transductionof LH-projecting BLA neurons
In order to control for stimulation of fibers of passage, as well as perform "phototagging"
during in vivo extracellular recordings, another group of animals underwent stereotaxic surgery
to transduce only LH-projecting BLA neurons with opsin. Using the same stereotaxic techniques
described above, I injected 1 pl of a retrograde virus, canine adenovirus 2 (CAV2) carrying crerecombinase, into the LH ( -.5 mm, + 1.1 mm ML, -5.1 mm DV) (Fig. IB). In some animals, I
injected a mixture (1:1) of CAV2-cre and AAV5 -CaMKIIa -mCherry to allow visualization of
injection site, given that CAV2-cre lacks a fluorophore. In the BLA, I infused an AAV 5
containing a 'double-floxed' inverted open reading frame (DIO) and hChR2(H134R)-eYFP
under a nonspecific neuronal promotor, the GTP-binding elongation factory family, EF- 1a. A
second group of animals received an infusion of virus with a control fluorophore (AAV5-EF1a15
DIO-eYFP). Given that only BLA neurons projecting to LH contain cre-recombinase with this
method, only the BLA to LH projection express ChR2-eYFP or control fluorophore, assuming
accurate infusion of virus in both regions. This method is referred to as "cre-DIO" hereafter.
With the same technique described previously, I implanted an optic fiber over the BLA to
permit optical stimulation of these neurons during behavioral assays. Animals were subjected to
behavioral assays 3-6 weeks following surgery, given the need for troubleshooting regarding
optimal timepoints to run animals with this technique.
Implantationof Chronic Optrodefor In Vivo ElectrophysiologicalRecordings
A subset of cre-DIO animals was reserved for in vivo extracellular electrophysiological
recordings and did not receive an optic fiber implantation following viral infusion. These animals
were maintained in the animal facility for 6-10 weeks following viral infusion, prior to
implantation of a chronic optrode, an electrode with an optic fiber (construction described
below), via a second surgery (Fig. IC).
During the second surgery, a craniotomy was drilled over the BLA for the optrode (-1.4
mm AP, +3.4 mm ML), and another smaller craniotomy over the posterior ipsilateral hemisphere
for the ground wire. Four skull screws were implanted around the optrode craniotomy. The
optrode was attached to a blue laser (473 nm, 20 mW) and the head-stage of the RZ5 recording
system (TuckerDavis Technologies, Alachua, FL, USA) to permit recording of electrical activity
while driving the optrode stereotaxically through the brain. The ground wire was secured at a
depth of - 1.5 mm in the second craniotomy, after the optrode was lowered to +1.5 mm DV. The
optrode was then driven down at approximately 0.01 mm/s until a depth of +4.7 mm DV, before
delivering optical stimulation (10 Hz, 5 ms pulse for 1 second(s)) to detect photoresponsive
units, in which units show time-locked firing in response to illumination. If such firing was
observed, constant illumination (1 s, every 30 s) was delivered to rule out a photoelectric effect,
16
in which electrical activity increases at light onset and offset, but does not represent a
photoresponsive unit. If the unit showed firing during the 1 s illumination (with possible tapering
due to blockade of sodium channels), the optrode was secured at that depth. If no
photoresponsive units appeared with optical stimulation, the optrode was lowered another 0.05
mm and illumination repeated. If no photoresponsive units were identified after reaching a
maximal depth of +5.15 mm DV, the optrode was retracted and re-implanted +.2mm AP from
previous coordinates. The same process was repeated and if no photoresponsive units were
found, the optrode was cemented (-1.6 mm AP, +3.4 mm ML, +5 mm DV) to avoid further
damage due to electrode tracks. The same cementing technique described previously was
performed, although more layers of cement were added and more time allotted between layers to
insure immobility of the implant. After the cement fully dried, the animal was disconnected from
the headstage and laser, and the incision was sutured. The animal was allowed to recover for 1-2
weeks with ad libitum food and water before behavioral assays and recordings began.
In vivo optogenetic manipulation
Virus construction andpackaging
The recombinant AAV vectors were serotyped with AAV 5 coat proteins and packaged by
the University of North Carolina Vector Core (Chapel Hill, NC). The maps of these constructs
are available online at www.optogenetics.org. The CAV2-cre virus was produced by Montpelier
Vectorology at BioCampus Montpelier (Montpelier, France).
Light delivery
During optogenetic behavioral assays, optical stimulation was delivered from a 100 mW
473 nm DPSS laser (OEM Laser Systems, Draper, UT). The laser was connected to a fiber-optic
rotary joint via a patch cord with FC/PC connectors on both ends (OEM Laser Systems, Draper,
UT), which was linked to another patch cord with a FC/PC connector on one end and a ferrule
17
with diameter matching the animal's implant on the other end. The mouse's implant was
attached to the patch cord via a ceramic mating sleeve (PFP, Milpitas, CA), and the rotary joint
allowed the animal to freely move without perturbation of optical stimulation. Light delivery was
modulated with a Master 8 pulse stimulator (A.M.P.I., Jerusalem, Israel) and triggered by
behavioral hardware (MedPC Associates, St. Albans, VT).
During in vivo extracellular recordings, the patch cord from the laser was attached to a
mechanical/optical commutator (Tucker-Davis Technologies), and light delivery was controlled
by the recording software (Tucker-Davis Technologies).
Behavioral assays
Animal Habituationto Handling
Two days prior to behavioral assays, I habituated animals to experimenter handling. In
each session, one per day for two days, cages were transported from the animal facility to the
laboratory behavioral rooms, and each animal was handled for 5 minutes. Handling consisted of
scruffing each animal, attaching the animal to a patch cord, allowing the animal to explore while
attached to the patch cord, and disconnecting the animal from the patch cord. After handling, the
animals were housed in darkness in the lab behavioral rooms for an hour before returning the
cages to the animal facility. An undergraduate, Melodi Anahtar, assisted with habituation of
animals.
Intracranialself-stimulation (ICSS)
4-5 weeks following surgery, animals were food restricted overnight and underwent
ICSS, an assay in which each animal was subjected to a self-stimulation session on two
consecutive days. For each session, animals were attached to a patch cord and allowed to freely
explore a dark, sound-proof conditioning chamber (MedPC Associates, St. Albans, VT)
equipped with auditory-stimulus generators and an infrared video camera, as well as an active
lx
and inactive nosepoke (Fig. 2A). Each session was initiated with the onset of low volume white
noise to mask any ambient noises. The active nosepoke triggered a tone and 1 second of optical
stimulation (473-nm, 20 Hz, 5-ms, 20 mW), and the inactive nosepoke triggered 1 second of a
distinct tone and no stimulation. Tones (1 and 1.5 kHz) and location of nosepokes were
counterbalanced across all animals. On the first day, both nosepokes were baited equally with
food (Fruit Loop), and, on the second day, the nosepokes were not baited. Access to food (4
hours) was given upon completion of the first day's session, and each animal was run at
approximately the same time each day. The apparatus was thoroughly cleaned with an acetic acid
solution between all animals to avoid bias, which an undergraduate, Melodi Anahtar, provided
assisted with. Med-PC software recorded timestamps for each nosepoke, and performance on the
second day was assessed using MATLAB and Microsoft Excel.
Open field test (OFT)
To assess effects of optogenetic manipulation upon locomotion and anxiety-related
behaviors, each animal underwent an OFT assay. Each animal was attached to a patch cable and
allowed to recover for 1-5 minutes prior to placement in an open field apparatus, a transparent
plexiglass box (50 x53 cm) (Fig. 2B). Within the software, the apparatus was divided into zones:
the center region (25 x 25 cm) and a border region (periphery). The 15-minute session consisted
of five alternating 3-minute epochs (OFF-ON-OFF-ON-OFF) with optical stimulation during ON
epochs (20 Hz, 5 ms, 10 mW). Laser power was reduced as compared to the ICSS assay due to
the longer duration of stimulation. During the behavioral assay, the experimenter remained
hidden behind an opaque screen, and the assay was conducted within a dimly lit room. A video
camera overhead recorded the animal's behavior, and an EthoVision XT video tracking system
(Noldus, Wageningen, Netherlands) acquired the mouse's location, velocity, and movement of
19
B
A
0 active nosepoke
t
counterbalanced
0 inactive nosepoke
.min *3mn 3min.3 min 3min
W -. . of
D
C
I min
3 min
Day 1
Order of light stim
counterbalanced
across animals
1 min
3 min
4
Day 2
Object location
counterbalanced
across animals
Object location
counterbalanced
across animals
E
0
nosepoke
sucrose port
Figure 2. Behavioral assays.
paired
A, In ICSS, animals are placed in a behavioral chamber that includes an active nosepoke,
optical
no
and
tone
second
a
to
paired
to a tone and optical stimulation, and an inactive nosepoke,
stimulation. Optical stimulation consists of a 1 s train of 20 Hz, 5 ms pulses (20 mW, 473 nm). B,
Animals explore an open field apparatus during 3-min alternating epochs of optical stimulation
with 20 Hz, 5 ms pulses (10 mW, 473 nm) and no stimulation. C, Animals are exposed to a
food pellet and novel object for 3-min on two consecutive days, with optical stimulation with 20
Hz, 5 ms pulses (10 mW, 473 nm) occurring on one of these days. D, Naive wildtype mice are
exposed to a pair of novel objects to insure no object preferences. E, The apparatus for the partial
reinforcement operant conditioning task consists of a nosepoke and sucrose port, with houselight
atop the sucrose port.
20
head, body, and tail. Melodi Anahtar helped compile the data from the epochs in Ethovision XT.
Microsoft Excel was used to further analyze this data.
Novelty & FeedingAssay
Animals were subjected to a novelty & feeding assay to determine effects of optical
stimulation upon novel object exploration and feeding behaviors. The novelty & feeding assay
consisted of two sessions, one per day, on consecutive days (Fig. 2C). Animals were attached to
a patch cable and allowed to recover for 1-5 minutes before placing the animal in its home cage
(all cagemates and the cage's nesting material was placed in another cage during the assay). The
home cage, as well as habituation, was used to reduce exploration of the environment and
promote interaction with the object and food, in accordance with previous findings regarding
object exploration and familiarity of environment (Besheer and Bevins, 2000). During each
session, the animal was allowed to habituate for 1 minute in the cage, followed by a 3 minute
exploration epoch. During the exploration epoch, a novel object and food pellet were placed at
opposite ends of the cage, away from the cage wall. Optical stimulation (20 Hz, 5 ms pulses, 10
mW) was delivered during the exploration epoch of one day's session and omitted during the
other day's session. The order of stimulation, order of objects, and location of objects were
counterbalanced across all animals, particularly as novel object assays can show an order effect,
with increased exploration on the first day as compared to the second day. Novel objects (plastic
princess figurine, constructed Lego figurine) were previously tested with a separate group of
animals, who exhibited no object preferences (described below). Novel objects were cleaned
with an ethanol solution between each animal and handled with new gloves. Food pellets,
identical to those normally used to feed the animals, were obtained from the animal facility, and
a new pellet, handled with new gloves, was used for each animal. For each animal, new gloves
21
were not allowed to contact any surface, object, or food item other than the given novel object or
food pellet prior to placement of the object in the cage.
Each session was recorded with a video camera overhead and analyzed by an observer
using ODLog behavioral analysis software (Macropod Inc., USA) to quantify object and food
interaction, digging, and eating behaviors. Object and food interaction (duration) was scored
such that proximity of the head within less than 2 cm of the object, in the direction of the object,
qualified as interaction. Rearing upon the object, or sitting near the object with the head directed
away, did not qualify as interaction, consistent with previously published methods (Antunes and
Biala, 2011).
Object Preference Assay
Object preferences among animals can create bias in novel object assays. While
individual preferences in experimental animals cannot be tested (the object would no longer be
novel in subsequent assays), the behavior of a separate group of wildtype mice can be used to
control for this possibility. Wildtype mice (n = 10) were placed in an enclosure, and two novel
objects (princess figurine, constructed Lego figurine) were deposited at either end, away from
the walls (Fig. 2D). The mice were allowed to freely explore for 10 minutes, and the video
recording was scored using ODLog software for interaction with each object, with the same
criteria described previously. The location of the objects was counterbalanced across all animals.
PartialReinforcement Sucrose Self-Administration Task
For in vivo extracellular recordings, an animal with an implanted optrode underwent an
appetitive operant conditioning task, in which each behavioral session was followed by a
"phototagging" session. Animals were food restricted for several days prior to the first training
session, and then attached to a headstage and patch cable (as described in previous sections). The
animal was placed in a dark, sound-proofed conditioning chamber (Med-PC Associates, Durban,
VT) which contained an auditory stimulus generator, infrared video camera, houselight,
nosepoke, and sucrose delivery port (30% sucrose solution in cage water) (Fig. 2E). Low volume
white noise was played throughout all sessions to mask ambient noise. In the first training
session, the nosepoke was baited (Food Loop), and the animal was allowed to freely explore the
apparatus. 50% of nosepokes triggered a 30-second tone (1 kHz), illumination of the houselight
(2.45 s) and delivery of a small volume of sucrose to the port (-.73 ml). If the animal did not
collect the sucrose, subsequent nosepokes triggered the tone without further sucrose delivery,
until the animal visited the port. For other nosepokes, no tone, houselight illumination, or sucrose
delivery was triggered, to control for motor activity in neural responses. The behavioral session
concluded once the animal collected 100 trials worth of sucrose, or an hour and a half had
passed. While still recording, the task was ended and a "phototagging" session was conducted to
identify any units as photoresponsive.
Visualizing viral transduction and fiber optic placement
Histology
Due to difficulties with seizures while the optimal timepoint for behavioral assays was
determined, animals did not undergo a c-fos stimulation protocol. Instead, mice were
anesthetized with pentobarbital sodium and transcardially perfused with ice-cold 4%
paraformaldehyde (PFA) in PBS (pH 7.3). Brains were extracted and fixed in 4% PFA overnight
and then transferred to 30% sucrose in PBS to equilibrate for 2-4 days. A sliding microtome
(HM430; Thermo Fisher Scientific, Waltham, MA) was used to collect 40 pm-thick coronal
sections, which were stored in PBS at 40 C. When immunohistochemistry was conducted,
sections were washed in Triton 0.3%/PBS and 3% normal donkey serum for one hour, followed
by incubation of sections with a DNA specific fluorescent probe (DAPI : 4',6-Diamidino-2Phenylindole (1:50,000)) for 1 hour with Triton 0.1%/PBS and 3% normal donkey serum at
-3
room temperature. Sections were then washed 4 times with PBS-IX and mounted on microscope
slides with PVD-DABCO. Melodi Anahtar assisted with a subset of these perfusions and some
sectioning of these brains.
Confocal microscopy
Confocal fluorescence images were collected using a 1OX/0.40NA or a 40X/1.30NA oil
immersion objective on an Olympus FV1000 confocal laser scanning microscope. Serial Z-stack
images were collected using the image analysis software (Fluoview, Olympus, Center Valley,
PA).Mice exhibiting eYFP somata expression in the central amygala, putamen, or piriform
cortex were excluded from analysis (n=7).
Statistical analysis
Statisical analyses were conducted using Microsoft Excel and Matlab software. The
threshold for significance for all results was p = .05 (denoted with *, p <.01 with **). All group
data are displayed as the mean
standard error of the mean (sem). Single variable differences
within within subject comparisons were found with two-tailed paired Student t-tests. Group
differences with two variables were analyzed with two-way ANOVA with Bonferroni post-hoc
tests.
In vivo extracellular electrophysiology
Optrode Construction
I constructed optrodes by attaching a 300 pm fiber (.37 NA) within a stainless steel
ferrule (1.25 mm diameter) to a 16-channel multielectrode array (Innovative Neurophysiology,
Durham, NC, USA). The ferrule was cemented to the electrode so that the optic fiber tip formed
an approximate 10 degree angle with the electrode tips, at a distance of 500 to 1500 ptm. The
optrode was constructed with this angle to insure that the light cone illuminated the electrode
tips.
24
BehavioralLearningCurves
To assess learning within the task, I employed a state space model that utilizes the
expectation maximization algorithm to estimate an individual animal's learning curve (the
probability of a correct response at each trial) (Smith et al., 2004). This state-space model
paradigm estimates learning at each trial, which enables precise comparisons between behavioral
changes and firing activity. In this paradigm, the learning criterion is defined by the confidence
limits of the learning curve; the learning trial occurs when the lower bound of the 90%
confidence interval remains above the probability expected by chance (50%) for the remainder of
the session. Analysis began by identifying each trial (initiated by a nosepoke) as correct (1) or
incorrect (0), with a correct response constituting sucrose retrieval within 10-s of tone onset, or
refraining from sucrose retrieval in the absence of a tone. A learning curve was generated from
this binary series by adapting MATLAB scripts, written by Anne Smith and obtained from the
Brown lab at MIT.
In Vivo ElectrophysiologicalRecordings and Phototaggingwith ChR2
Following each day's behavioral session, phototagging with 473 nm laser (30-40 mW)
commenced within the same recording session, in order to identify photoresponsive units. The
phototagging session consisted of pseudorandomly dispersed optical stimulation of 1 s constant
light or 10 s of 1 Hz light (5 ms pulses), with 10 repetitions of each stimulation type.
Electrophysiological data and behavioral timestamps were exported from the TDT
system. Plexon offline sorter was used to sort waveforms with principal component analysis, and
further analysis was conducted with NeuroExplorer and MATLAB.
Visualizing optrodeplacement
Prior to sacrificing an animal implanted with an optrode, electrolytic lesions were created
%
to localize electrode tips. After anesthesization with isoflurane (5 % for induction, 1.5-2.0
after), a 19.6 mA current (15 s) was passed through each channel with sorted units. After
sufficient time for gliosis passed (30 minutes), the animals were anesthesized with pentobarbital
and perfused with the same technique described previously.
26
Results
Opticalstimulation of BLA fibers in LH supports ICSS
To test whether optogenetic activation of BLA terminals in LH supports ICSS, I
expressed Channelrhodopsin-2 (ChR2)-eYFP fusion protein in BLA pyramidal neurons in
experimental animals and eYFP in control animals matched for age, incubation time, and
illumination parameters. Prior to the behavioral assay, I implanted an optic fiber over the LH, in
the hemisphere ipsilateral to the viral infusion. Following all behaviors, confocal imaging of
tissue sections revealed viral transduction of BLA neurons without leakage into adjacent areas
and correct placement of the optic fiber (ChR2, n = 8; eYFP, n = 7) (Fig. 3). The data from all
animals that showed viral leakage or incorrect fiber placement were excluded from primary
analysis (n = 7). The surgeries for two ChR2 animals, one of which was excluded due to viral
leak, were performed by Chris Leppla.
ChR2 animals that underwent the ICSS assay showed a robust preference for the active
over the inactive nosepoke during the second, unbaited session (Fig. 4A). The difference in
nosepokes between active and inactive nosepokes among ChR2 animals was statistically
significant as compared to eYFP (p < .0 1) (Fig. 4B). A representative trace showing a ChR2
animal's performance over the course of a session is shown (Fig. 4C). Of note, animals that
showed viral leak into CEA did not exhibit a preference for either nosepoke, despite ample
expression in the BLA.
ICSS most likely arisesfrom reward-relatedprocesses
To determine the effects of optical stimulation upon locomotion or anxiety-related
behaviors, I conducted an OFT assay, which consisted of five alternating 3-minute epochs of
optical stimulation (OFF-ON-OFF-ON-OFF). A subset of ChR2 animals (n = 3) were excluded
from analysis due to seizures during this assay, even with minimal illumination, and were not
27
Figure 3. Confocal imaging reveals viral transduction of BLA and fiber placement over LH.
A, A confocal image shows ChR2-e YFP (green) expression in the BLA. B, A confocal image
displays optic fiber placement (red) over the LH with fibers expressing ChR2-eYFP (green). C, A
confocal image from another animal shows a magnified view of correct fiber placement over the
LH. D, An example of viral leakage beyond the BLA, into the CEA, which resulted in exclusion
of this animal from primary analysis.
28
A 600
.ChR2
eYFP
**
B
300
(n = 8)
I-,
C
4-
0
0. 300
a)
a)
0
z
a)
a)
0
Active
C
ChR2 eYFP
Inactive
450
0225
a)
LA.
0
z
Inactive
Time(min)
60
Figure 4. Optical stimulation of BLA fibers supports ICSS.
nosepoke
A, ChR2 animals (blue, n = 8) exhibited a robust preference for the active over inactive
B, ChR2
7).
=
n
(red,
controls
eYFP
to
compared
during the second, unbaited session of ICSS, as
and
active
in
difference
mean
animals' preference for the active nosepoke, as reflected by the
C,
1).
.0
<
(p
controls
inactive nosepokes, was highly statistically significant as compared to eYFP
unbaited session of
A representative trace displaying a ChR2 animal's performance in the second,
session.
the
of
ICSS, with active nosepokes (blue) increasing over the course
29
subjected to any further assays. Only eYFP controls run during the same session as non-seizing
ChR2 animals were retained in further analysis. Of the remaining ChR2 animals (n = 5), there
was no substantial change in locomotion as compared to eYFP controls, reflected by similar
mean distances moved and velocities across epochs (Fig 5A,B). ChR2 animals also exhibited
similar anxiety-related behaviors to eYFP controls, with similar mean duration in the center and
border zones of the apparatus (Fig. 6A). For the second epoch of optical stimulation, ChR2
animals exhibited an increase in time spent in the center zone, but this difference was not
significant. A representative ChR2 animal's movements within the apparatus' zones are shown
(Fig. 6B).
Novel object exploration & feeding behavior in a subset of ChR2 animals
Given the implication of the BLA to LH projection in CS-potentiated feeding, the
remaining ChR2 animals (n = 2) were subjected to a novelty & feeding assay. A novel object
was included in this assay to reveal any effects upon exploration and anxiety-related behaviors.
ChR2 animals displayed similar levels of novel object exploration with and without optical
stimulation (Fig. 7A), as well as similar exploration of food with and without optical stimulation
(Fig. 7B). ChR2 animals showed a lower interaction difference as compared to eYFPs, with the
interaction difference calculated by subtracting the duration of interaction with the food pellet
from interaction with the novel object (Fig. 7C). Experimental animals displayed no substantial
differences in digging (Fig. 7D), and optical stimulation did not elicit any feeding behavior in
either experimental or control animals. As a control, wildtype animals were subjected to a
separate assay to test for object preferences, but there was no difference in interaction time with
each of the novel objects (Fig. 8A).
Absence of optical self-stimulation with putative transduction ofLH-projectingBLA neurons
The ICSS observed with viral transduction of BLA neurons may arise from stimulation of
30
A
2000
E 1600
0
E 1200
-o-ChR2
C
-o-eYFP
4-800
400
0
B
OFF
ON
OFF
ON
OFF
12
10
8
E
6
4
2
0
OFF
ON
OFF
ON
OFF
Figure 5. ICSS in ChR2 animals does not arise from a locomotor effect.
A, ChR2 animals (blue, n =5) showed no difference in mean distances moved ( sem, cm) during
ON and OFF epochs, or as compared to eYFP controls (red, n =5). B, ChR2 animals exhibited
no changes in mean velocity ( sem, cm/s) across epochs, or as compared to eYFP controls. ON
epochs delivered 20 Hz, 5 ms pulses (10 mW, 473 nm).
31
A
25
20
15
. 10
a)
E
5
0
OFF
ON
OFF
OFF
ON
B
OFF
ON
Figure 6. Anxiety-related behaviors do not appear to contribute to ICSS in ChR2 animals.
spent
A, ChR2 animals (blue, n =5) did not show a significant change in mean duration ( sem)
ChR2
in the center zone as compared to eYFP controls (red, n = 5). B, A representative track of a
zones
center
and
border
through
epochs
(blue)
animal's movement during OFF (black) and ON
(dashed line).
32
B
A
20
20
ChR2
eYFP
D
C
0
(U 10M
10
4-J-
00
0
0
ON
OFF
ON
OFF
O
F
100
D
C
20
10
4-J
0
0
ON
OFF
-4
(
Figure 7. Preliminary results of novel object exploration and feeding with optical stimulation.
A, ChR2 animals (blue) did not show a signficant difference in mean novel object interaction
sem, s) or food interaction (B), as compared to eYFP animals (red) with optical stimulation of 20
Hz, 5 ms pulses (10 mW, 473 nm). C, The interaction difference, reflecting the difference between
novel object and food interaction, was not significantly different with optical stimulation among
ChR2 animals, although lower than with eYFP controls. D, ChR2 animals displayed no substantial
change in time spent digging with optical stimulation or no stimulation relative to eYFP animals.
33
20
0
10
4-0
0
0
Object A
(Princess)
Object B
(Lego)
Figure 8. Absence of object preference among naive wildtype animals. In a control assay to
rule out object preferences among animals, wildtype mice (n =10) showed no significant difference
in interaction time ( sem, s) with either novel object used in the novelty & feeding assay.
34
fibers of passage, rather than terminals in LH, particularly since the BLA possesses efferent
connections to the ventromedial hypothalamus (VMH) (Petrovich et al., 2001). To control for
this possibility, I used a cre-DIO method to selectively transduce LH-projecting BLA neurons
with ChR2 and implanted an optic fiber over the BLA ipsilateral to viral injections. These
animals underwent the ICSS assay at four and five weeks following surgery, but did not exhibit a
preference for the active over inactive nosepoke at either timepoints (Fig. 9A). They also showed
no difference in locomotion or anxiety-related behaviors as compared to eYFP controls (Fig.
9B). However, histological confirmation of viral transduction and fiber placement was not
conducted.
Feasibilityof in vivo electrophysiologicalrecordings during operantappetitive conditioning
While optical self-stimulation observed with CamKlIa-ChR2 animals may indicate that
this projection supports motivated behavioral responding, the actual encoding properties of these
neurons in the context of appetitive learning remains unclear. In order to characterize the firing
activity of LH-projecting BLA neurons, I selectively transduced these neurons with ChR2 using
the cre-DIO method in order to permit phototagging of units during electrophysiological
recordings (n = 2). During optrode implantation, I observed short latency (< 8ms), time-locked
firing in response to optical stimulation (1 s constant & 10 Hz, 5 ms pulses) in one animal (Fig.
1 OA). After behavior, confocal imaging confirmed viral transduction of BLA neurons and
electrode localization in the BLA (Fig. lOB). An example of sorted units, while not photoresponsive, from a behavioral session with the partial reinforcement operant conditioning task is
shown (Fig. I0C).
Although photoresponsive units were not present during subsequent recording sessions,
the animal was able to acquire an operant appetitive conditioning task while attached to the
headstage and patch cable, which can sometimes interfere with the animals' ability to perform. A
_)5
.
.........
...
B
A
70
60
* ChR2
* eYFP
50
MC 0
tA
'9
0
T
-5
40
CL
4)
4A
0
(U 10
Z 30
20
S15
I
20
-
20
U
25
10
0
Inactive
Active
D
C
2000
E
.4
25
20
1600
15
0 1200
E
a)
C
a)
10
C
a)
5
U
800
E
1=
400
0
-5
0
OFF
ON
OFF
ON
OFF
ON
OFF
ON
OFF
ON
-10
Figure 9. No significant effect of optical stimulation of putative LH-projecting BLA neurons.
A, ChR2 (blue) and eYFP (red) animals did not show a preference for the active over inactive
nosepoke during ICSS. B, The mean difference in active and inactive nosepokes ( , sem) was not
statistically significant among ChR2 (n = 6) animals, as compared to eYFP controls (n = 2). C,
ChR2 animals (blue) exhibited no substantial change in mean distance moved ( sem, cm) across
epochs or in comparison to eYFP controls (red). D, ChR2 animals hsowed no substantial difference in mean time spent in the center zone ( sem, s).
36
Ii I
A
11
111
11
I
11
II
I
1111 11
I I
1 11 l l
II
11
II 1 1
II
I I I I
II
111 1111 111 111 :1 I
I
I
I II
ll 11
11 11
I
a:
O'\
c
·;:::::
u:::
11
IILI l ~l
-1
1
1 11
1 1
11
I
1
I
I I
I
"N so
~
+-'
tO
11111
II
Ill
Q)
Ii
1 11 1
11
1 1
I
"Nso
~
J
1 1
1 11 1
1111
II I
1
1
'~k-
0
Time (sec)
I
Q)
+-'
tO
I ll~
a:
O'\
c
·;:::::
u:::
_JI J1jiJ
u
·-
0
Time (sec)
c
Figure 10. Successful transduction of LB-projecting BLA neurons with ChR2-eYFP and in
vivo extracellular recordings. A, During optrode implantation, a unit shows short latency (<
8ms), time-locked firing in response to 1 s constant photostimulation (20 mW, 473 mn) and 10
Hz photostimulation for 1 s (5 ms pulses). B, A confocal image shows an electrolytic lesion on
the border of the BLA, with viral expression of ChR2-e YFP (green) limited to the BLA. C ,
Representative clusters show cell sorting through principal component analysis and corresponding
waveforms (D), reflecting neural activity recorded during the animal 's acquisition of the task.
37
learning curve from the animal's second recording session displays the learning trial (Fig. 11).
This curve was generated using code that I adapted from scripts written by Anne Smith, which
are accessible through the Brown Lab website. Previously published electrophysiological
recordings with this task in mice had only been conducted after the animal learned the task,
which precluded analysis of acquisition (Nieh et al., 2015).
31
T
0
00.
0
~0
0
I
I
I
I
50
I
I
A
100
Trial Number
Figure 11. Behavioral learning curve displays animal's acquisition of the task.
The learning curve (red) shows an animal's performance on the second session of a partial
reinforcement operant conditioning task, with incorrect trials (dark grey) in the beginning and
correct trials (light grey) increasing over the session. The learning trial (green arrows) occurs when
the lower bound of the 90% CI (blue lines) remains above the probability expected by chance
(blasck line) for the rest of the session.
39
Discussion
Projectionfrom BLA to LH may support motivated behavioralresponding
Through optogenetic techniques, I have shown that the projection from BLA to LH may
support optical self-stimulation, which has not been previously investigated. Previous methods
did not permit selective, temporally precise manipulation of LH-projecting BLA neurons, as
most studies have relied upon lesions, which allow for adaptation and effects upon fibers of
passage.With viral transduction of BLA neurons and a fiber over the LH, ChR2 animals
demonstrated a strong preference for the active over inactive nosepoke.
While the animals' optical self-stimulation indicates a possible role for this projection in
motivated behavioral responding, this result could also reflect non-reward related phenomenon.
For example, an effect of optical stimulation upon locomotion or behavioral flexibity could
prevent the animal from moving away from the active nosepoke. However, a subset of ChR2
animals did not display any difference in locomotion, as compared to eYFP controls, in the OFT
assay. Optical stimulation could also prevent the animal from switching its current strategy,
either by diminishing exploration or promoting perserveration. However, no difference in center
zone exploration appeared during the OFT assay in experimental animals as compared to
controls. Although ChR2 animals exhibited a lower baseline of novel objection exploration than
eYFPs, in the novelty & feeding assay, the low number of animals precludes definitive
conclusions.
Given the absence of exploratory effects, the optical self-stimulation observed most likely
stemmed from reward-related processes, although addition assays would enrich interpretation of
the ChR2 animals' preference. If stimulation promotes perseverative behaviors, absent of any
appetitive influence, a conditioned place preference (CPP) assay would likely reveal this effect.
CPP is a three day assay that requires a three-chamber apparatus, in which one chamber is linked
40
to optical stimulation and another to no stimulation. The animal is exposed to each of these
chambers, one per day, and on the third day allowed to freely explore the apparatus, with optical
stimulation withheld. The animals' preference for the stimulated chamber on the third day,
reflected by increased time within that chamber, is thought to indicate an appetitive effect of
stimulation, while avoidance implies an aversive effect. If optical stimulation of BLA to LH
prevents the animal from switching strategies, but has no appetitive effect, animals will exhibit
no preference for the stimulated chamber, and if aversive, will seek out the nonstimulated
chamber. In contrast, in a real-time place preference (RTPP) assay, such animals might still show
a preference for optical stimulation, due to an inability to leave the chamber paired to simulation.
In RTPP, a single day assay, animals are allowed to freely explore all chambers of the apparatus,
in which one chamber triggers optical stimulation. Thus, the CPP assay could help substantiate
this projection's role in reward-related behavior, although it would not differentiate incentive
salience (wanting) versus hedonic impact (liking) (Berridge et al., 2009). Assays with analysis
for facial 'liking' expressions could help distinguish these psychological processes, although
studies of CS-potentiated feeding assays and LH (Holland et al., 2002, 2002; Petrovich et al.,
2005) indicate a more likely influence upon incentive salience than hedonic impact.
Analysis of excluded animals indicatespossible opposing effect of CEA to LHprojection
Of note, a subset of animals that were excluded from primary analysis due to viral
leakage did not exhibit ICSS, despite ample transduction of BLA neurons. This absence of ICSS
appeared to occur when virus leaked into the CEA, but not the putamen, raising the possibility
that the CEA to LH projection opposes the effect of the BLA to LH projection. In contrast to
BLA, CEA contains long range GABAergic projections (Sah et al., 2003), such that the CEA
may directly inhibit the neurons that BLA synapses onto, or the CEA may indirectly block ICSS
through inhibition of another subset of LH neurons. If this result persists with more animals, ex
41
vivo intracellular recordings in slice may be warranted to untangle the circuit dynamics,
particularly if opsins with non-overlapping wavelengths (CIVI and ChR2) (Yizhar et al., 2011)
could be used to activate each projection separately in the same animal while recording from LH
neurons.
Activation of BLA fibers in LH do not nonspecifically triggerfeeding
Animals in familiar environments are known to preferentially interact with novel objects
as opposed to familiar objects, which has suggested an appetitive aspect to novelty exploration
(Bevins et al., 2002). Thus, given the implication of LH in feeding and reward, I conducted a
modified novel object assay. To test for possible effects upon reward processing, feeding, and
anxiety-related behaviors, I optically stimulated BLA terminals in the LH during exposure to a
novel object and food pellet. ChR2 animals did not exhibit an appreciable difference in novel
object exploration or food interaction with and without optical stimulation. However,
experimental animals did show a low baseline exploration level, also reflected in the low
difference interaction score, especially as compared to eYFP animals. Most likely, this low
baseline is an artifact of the low number of animals, perhaps with the eYFP baseline inordinately
raised by the performance of a single animal. A true effect of optical stimulation on ChR2
animal's baseline appears unlikely, given a reduction in center zone exploration was not
observed during the OFT, which reflected more animals. Other possibilities include an effect of
illumination upon memory, with stimulation on the first day affecting behavior on the second,
but counterbalancing of order would most likely dilute this effect. The order of assays could also
affect baseline exploration, with optical stimulation in previous assays (OFT and ICSS)
producing plasticity within the circuit, which may warrant separate groups of naive experimental
animals in future experiments.
42
Familiar food was employed in this assay instead of a familiar object, the main-stay of
novel object assays, in order to gauge consummatory behaviors. Optical stimulation of
GABAergic VTA-projecting LH neurons is known to trigger gnawing behavior in a very
pronounced fashion (Nieh et al., 2015), and by including both an object and food, I could assay
the occurrence of any feeding behavior and whether gnawing motions were food specific.
Optical stimulation did not trigger any gnawing or feeding behavior in this assay or any others,
which indicates that BLA fibers in LH most likely do not solely project on to GABAergic VTAprojecting LH neurons, given that BLA projections are glutamatergic. (LH-projecting BLA
neurons could still potentially project to this population, with synapses onto other neurons
eliminating gnawing behavior.)
This absence of feeding behavior also enriches interpretation of previous studies of CSpotentiated feeding. Even though lesions did not affect baseline food consumption, it was unclear
whether BLA to LH simply triggers feeding in any circumstance, or performs another function,
such as endowing the CS with reinforcing properties, thus causing the animal to pursue the US
despite energy-related signals indicating satiation (Baxter and Murray, 2002; Holland et al.,
2002; Petrovich et al., 2002). While only a few mice, the lack of any feeding behavior supports
the idea that LH-projecting BLA neurons help a CS acquire reinforcing properties, rather than
nonspecifically triggering feeding.
LH anatomy necessitates more specific optogenetic methods
Although transduction of the entire BLA and implantation of an optic fiber over terminals
has permitted selective manipulation of projections in previous studies (Felix-Ortiz and Tye,
2014; Felix-Ortiz et al., 2013; Stuber et al., 2011), the anatomy of the BLA to LH projection
most likely requires a more specific method. In addition to LH, the BLA possesses a substantial
projection to the VMH (Petrovich et al., 2001), such that an optic fiber over LH may activate
43
fibers terminating in VMH. As a control for such an effect, many studies conduct optical
stimulation prior to perfusion and stain for immediate early genes, such as c-fos (Lammel et al.,
2012; Tye et al., 2011). However, the LH sends a small projection to VMH (Ter Horst and
Luiten, 1987), such that c-fos expression in VMH could represent activation directly via ChR2,
or secondary activation by LH neurons responding to illumination. A glutamate receptor
antagonist could be infused into the LH prior to ICSS, and any resulting absence of optical selfstimulation would strongly substantiate the role of LH-projecting BLA neurons in motivated
behavioral responding. However, the close proximity of the VMH, immediately adjacent to the
LH may obfuscate conclusions, and a dye infused along with the antagonist may not spread in
the same manner as the drug or persist in the tissue until perfusion, preventing confirmation of
glutamate blockade in the LH alone.
The ChR2 animals exhibited a very short window in which optical stimulation supported
ICSS but did not provoke seizures, highlighting the need for controls for possible antidromic
activation. It is unclear whether LH-projecting BLA neurons possess collaterals, or where such
collaterals terminate, but infusion of lidocaine into the BLA during ICSS would block such an
effect, while still permitting optical activation of terminals in LH. However, given the technical
difficulties concerning fibers of passage, discussed above, I proceeded to targetting the
projection with another method rather than attempting pharmacological inactivation of BLA cell
bodies.
Preliminaryresults with cre-DIO method
Given the aforementioned confounds, I tried to selectively transduce only LH-projecting
BLA neurons using a cre-DIO method. Previous retrograde viruses, such as HSV-cre, produced
minimal penetrance in our lab, even after lengthy incubation times (6 months or more), while
quicker methods, involving a pseudorabies virus, often produced toxicity before all control
44
assays could be completed. As a result, I opted to infuse a CAV2-Cre virus into the LH, which,
as shown by the confocal imaging of an optrode animal, resulted in expression of ChR2 in a
subset of BLA neurons. However, the absence of a fluorophore in this virus still creates
difficulties due to the proximity of VMH, so I simultaneously infused a virus bearing a
fluorophore alone, under the CamKII promotor. A second virus may still spread differently than
the CAV2-cre virus, although it would most likely produce a closer approximation than dye and
persist in the tissue long enough for histological examination. In future experiments, a
fluorophore under an even less specific neuronal promotor may also be preferable.
With this method and an optic fiber implanted over the BLA, I did not observe optical
self-stimulation in the ICSS assay, however, several limitations of this study mean that LHprojecting BLA neurons may still promote motivated behavioral responding. First and foremost,
incorrect placement of the viruses or the optic fiber could explain the absence of an effect. The
previous method may have also targetted a different subset of LH-projecting BLA neurons than
the cre-DIO method, as a single optic fiber cannot illuminate the entire LH, given its lengthiness.
The virus may have spread further than the light cone in previous experiments, and BLA neurons
projecting to different parts of LH may have opposing effects, resulting in an absence of a result
with ICSS. Repetition of the experiment with infusion of a smaller volume of virus may be
warranted, and dissection of differing effects depending on the projection target along LH could
be fruitful. If further experiments with the cre-DIO method show optical self-stimulation, it is
likely that LH-projecting BLA neurons support motivated behavioral responding.
Future directionsfor investigating the encodingproperties of LH-projectingBLA neurons
In order to characterize the encoding properties of this subset of neurons, I selectively
transduced LH-projecting BLA neurons with the cre-DIO method and implanted an optrode in
the BLA, so that LH-projecting BLA neurons could be identified via their response to
45
illumination. After recovery, the animal was subjected to a partial reinforcement operant
conditioning task, which was used due to previous demonstration of deficits with this task
following NMDA blockade of the BLA (Tye et al., 2008), making the BLA's involvement likely.
I also employed this task rather than a Pavlovian task because the animal's initiation of each trial
helps control for attentional effects.
During surgery, I recorded from photoresponsive units, but these units did not persist in
behavioral recording sessions. However, an animal successfully learned the task while connected
to a headstage and patch cable, as evidenced by the learning curve, which is taken from the
second day of recording. The animal's performance begins above 50% (that expected by chance)
most likely due to learning during the first day's session. The identification of photoresponsive
units during surgery also improves upon previous methods with HSV-Cre in our lab, which had
produced such sparse labelling in the BLA that few photoresponsive units appeared in
recordings. Thus, while I did not record phototagged units during the actual task, I did
demonstrate the feasability of this experiment, which merits further investigation given my
preliminary results with ICSS and this projection.
Possiblefiringpatterns among LH-projectingBLA neurons during an appetitive task
LH-projecting BLA neurons may exhibit several forms of encoding in the context of
appetitive learning. Previous electrophysiological recordings during appetitive conditioning and
operant tasks have demonstrated pronounced diversity in the firing activity of BLA neurons, with
responses to rewards (unconditioned stimuli (US)), reward-predictive cues (CS) (Muramoto et
al., 1993; Paton et al., 2006; Sanghera et al., 1979; Tye and Janak, 2007; Tye et al., 2008; Uwano
et al., 1995), expectation of reward (Schoenbaum et al., 1998), and omission of expected reward
(Roesch et al., 2010; Tye et al., 2010). The effect of contralateral BLA and LH lesions on CSpotentiated feeding indicates that LH-projecting BLA neurons may evoke the value of the CS
46
(Petrovich and Gallagher, 2003). Other electrophysiological studies in the primate and rat have
suggested that BLA neurons encode current stimulus-value associations (Baxter and Murray,
2002; Belova et al., 2008; Paton et al., 2006; Schoenbaum et al., 2007) rendering it likely that
LH-projecting BLA neurons will respond to reward-predictive cues in an appetitive conditioning
task.
If LH-projecting BLA neurons encode the value of reward-predictive cues, units may
display a phasic response to the US during initial recordings, followed by response to the CS and
US as the animal learns, with depolarization due to the US allowing for NMDA-dependent
plasticity. After acquisition, such neurons may decrease firing over the course of the session, as
the value of the outcome diminishes due to satiation. LH-projecting neurons may also respond to
the presence or absence of sucrose, regardless of expectation, as found in previous studies of
BLA neurons (Tye et al., 2010), or they may represent motor activity related to conditioned
approach. If such activity is observed, recordings in animals with unpaired sucrose and tones will
potentially help differentiate activity related to movement execution and reward expectation.
After acquisition, recordings could also occur while altering the animal's learned
associations, such as delivering greater sucrose than expected or omitting expected sucrose. Such
manipulations would allow for further analysis of value coding and signed and unsigned reward
prediction error (RPE). If LH-projecting BLA neurons exhibit value coding, then the response
will diminish across blocks in which the tone no longer predicts surose delivery, and possibly
increase with greater sucrose. If these neurons encode signed reward prediction error (RPE),
units will exhibit inhibition to sucrose omission and an increase in firing to greater sucrose
delivery (Schultz, 2006). During blocks, this response will diminish as the outcome becomes
expected, with perhaps greater firing transitioning to the cue in blocks involving greater sucrose
delivery. Rather than signed prediction error, previous recordings have reported a subset of BLA
47,
neurons that appear to encode unsigned prediction error, in which an unexpected outcome
increases firing, regardless of valence (Roesch et al., 2010). If LH-projecting BLA neurons
encode unsigned prediction error, the omission of sucrose or delivery of greater sucrose will both
increase firing, and this firing will diminish across the block. Furthermore, LH-projecting BLA
neurons may all show similar firing properties, as reported with VTA-projecting LH neurons
(Nieh et al., 2015), or they may show heterogeneity across neurons or a combination of encoding
properties within a single unit.
Conclusion
Although both the BLA and LH have been implicated in reward processing, pre-existing
techniques have rendered it difficult to dissect the function of this projection in appetitive
operant conditioning. Through optogenetic manipulations, I have found preliminary evidence
that LH-projecting BLA neurons support motivated behavioral responding, and I have
demonstrated the use of in vivo extracellular recordings with identification of these neurons via
their photoresponse. Through a more specific method of manipulating the BLA to LH projection,
future experiments could substantiate these preliminary findings, as well as delve into the
encoding properties of this subset of BLA neurons.
49
References
Adamantidis, A.R., Zhang, F., Aravanis, A.M., Deisseroth, K., and Lecea, L. de (2007). Neural
substrates of awakening probed with optogenetic control of hypocretin neurons. Nature 450,
420-424.
Anand, B.K., and Brobeck, J.R. (1951). Localization of a "feeding center" in the hypothalamus
of the rat. Proc. Soc. Exp. Biol. Med. Soc. Exp. Biol. Med. N. Y. N 77, 323-324.
Antunes, M., and Biala, G. (2011). The novel object recognition memory: neurobiology, test
procedure, and its modifications. Cogn. Process. 13, 93-110.
Barone, F.C., Wayner, M.J., Scharoun, S.L., Guevara-Aguilar, R., and Aguilar-Baturoni, H.U.
(1981). Afferent connections to the lateral hypothalamus: A horseradish peroxidase study in the
rat. Brain Res. Bull. 7, 75-88.
Baxter, M.G., and Murray, E.A. (2002). The amygdala and reward. Nat. Rev. Neurosci. 3, 563-
573.
Beckstead, R.M., Domesick, V.B., and Nauta, W.J.H. (1993). Efferent Connections of the
Substantia Nigra and Ventral Tegmental Area in the Rat. In Neuroanatomy, (Birkhauser Boston),
pp.449-475.
Belova, M.A., Paton, J.J., and Salzman, C.D. (2008). Moment-to-Moment Tracking of State
Value in the Amygdala. J. Neurosci. 28, 10023-10030.
Berridge, K.C., Robinson, T.E., and Aldridge, J.W. (2009). Dissecting components of reward:
"liking", "wanting", and learning. Curr. Opin. Pharmacol. 9, 65-73.
Berthoud, H.-R., and Munzberg, H. (2011). The lateral hypothalamus as integrator of metabolic
and environmental needs: From electrical self-stimulation to opto-genetics. Physiol. Behav. 104,
29-39.
Besheer, J., and Bevins, R.A. (2000). The role of environmental familiarization in novel-object
preference. Behav. Processes 50, 19-29.
Bevins, R.A., Besheer, J., Palmatier, M.I., Jensen, H.C., Pickett, K.S., and Eurek, S. (2002).
Novel-object place conditioning: behavioral and dopaminergic processes in expression of
novelty reward. Behav. Brain Res. 129, 41-50.
Bielajew, C., and Shizgal, P. (1986). Evidence implicating descending fibers in self-stimulation
of the medial forebrain bundle. J. Neurosci. 6, 919-929.
Brobeck, J.R. (1946). Mechanism of the Development of Obesity in Animals with Hypothalamic
Lesions. Physiol. Rev. 26, 541-559.
Bums, L.H., Everitt, B.J., and Robbins, T.W. (1994). Intra-amygdala infusion of the N-methyl-daspartate receptor antagonist AP5 impairs acquisition but not performance of discriminated
approach to an appetitive CS. Behav. Neural Biol. 61, 242-250.
Caine, S., Heinrichs, S.C., Coffin, V.L., and Koob, G.F. (1995). Effects of the dopamine D-1
antagonist SCH 23390 microinjected into the accumbens, amygdala or striatum on cocaine selfadministration in the rat. Brain Res. 692, 47-56.
Ciocchi, S., Herry, C., Grenier, F., Wolff, S.B.E., Letzkus, J.J., Vlachos, I., Ehrlich, I., Sprengel,
R., Deisseroth, K., Stadler, M.B., et al. (2010). Encoding of conditioned fear in central amygdala
inhibitory circuits. Nature 468, 277-282.
DeFalco, J., Tomishima, M., Liu, H., Zhao, C., Cai, X., Marth, J.D., Enquist, L., and Friedman,
J.M. (2001). Virus-Assisted Mapping of Neural Inputs to a Feeding Center in the Hypothalamus.
Science 291, 2608-2613.
Drevets, W.C. (2001). Neuroimaging and neuropathological studies of depression: implications
for the cognitive-emotional features of mood disorders. Curr. Opin. Neurobiol. 11, 240-249.
Felix-Ortiz, A.C., and Tye, K.M. (2014). Amygdala Inputs to the Ventral Hippocampus
Bidirectionally Modulate Social Behavior. J. Neurosci. 34, 586-595.
Felix-Ortiz, A.C., Beyeler, A., Seo, C., Leppla, C.A., Wildes, C.P., and Tye, K.M. (2013). BLA
to vHPC inputs modulate anxiety-related behaviors. Neuron 79, 658-664.
Gutierrez, R., Lobo, M.K., Zhang, F., and de Lecea, L. (2011). Neural integration of reward,
arousal, and feeding: Recruitment of VTA, lateral hypothalamus, and ventral striatal neurons.
IUBMB Life 63, 824-830.
Harris, G.C., Wimmer, M., and Aston-Jones, G. (2005). A role for lateral hypothalamic orexin
neurons in reward seeking. Nature 437, 556-559.
Hatfield, T., Han, J.S., Conley, M., Gallagher, M., and Holland, P. (1996). Neurotoxic lesions of
basolateral, but not central, amygdala interfere with Pavlovian second-order conditioning and
reinforcer devaluation effects. J. Neurosci. 16, 5256-5265.
Haubensak, W., Kunwar, P.S., Cai, H., Ciocchi, S., Wall, N.R., Ponnusamy, R., Biag, J., Dong,
H.-W., Deisseroth, K., Callaway, E.M., et al. (2010). Genetic dissection of an amygdala
microcircuit that gates conditioned fear. Nature 468, 270-276.
Hoebel, B.G., and Teitelbaum, P. (1962). Hypothalamic control of feeding and self-stimulation.
Science 135, 375-377.
Holland, P.C., Hatfield, T., and Gallagher, M. (2001). Rats with basolateral amygdala lesions
show normal increases in conditioned stimulus processing but reduced conditioned potentiation
of eating. Behav. Neurosci. 115, 945-950.
51
Holland, P.C., Petrovich, G.D., and Gallagher, M. (2002). The effects of amygdala lesions on
conditioned stimulus-potentiated eating in rats. Physiol. Behav. 76, 117-129.
Hurd, Y.L., Weiss, F., Koob, G.F., And, N.-E., and Ungerstedt, U. (1989). Cocaine
reinforcement and extracellular dopamine overflow in rat nucleus accumbens: an in vivo
microdialysis study. Brain Res. 498, 199-203.
Jimenez, S.A., and Maren, S. (2009). Nuclear disconnection within the amygdala reveals a direct
pathway to fear. Learn. Mem. 16, 766-768.
Kempadoo, K.A., Tourino, C., Cho, S.L., Magnani, F., Leinninger, G.-M., Stuber, G.D., Zhang,
F., Myers, M.G., Deisseroth, K., Lecea, L. de, et al. (2013). Hypothalamic Neurotensin
Projections Promote Reward by Enhancing Glutamate Transmission in the VTA. J. Neurosci. 33,
7618-7626.
Lammel, S., Lim, B.K., Ran, C., Huang, K.W., Betley, M.J., Tye, K.M., Deisseroth, K., and
Malenka, R.C. (2012). Input-specific control of reward and aversion in the ventral tegmental
area. Nature 491, 212-217.
LeDoux, J.E., Cicchetti, P., Xagoraris, A., and Romanski, L.M. (1990). The lateral amygdaloid
nucleus: sensory interface of the amygdala in fear conditioning. J. Neurosci. 10, 1062-1069.
Mailkovai, L., Gaffan, D., and Murray, E.A. (1997). Excitotoxic Lesions of the Amygdala Fail to
Produce Impairment in Visual Learning for Auditory Secondary Reinforcement But Interfere
with Reinforcer Devaluation Effects in Rhesus Monkeys. J. Neurosci. 17, 6011-6020.
Maren, S. (2005). Synaptic mechanisms of associative memory in the amygdala. Neuron 47,
783-786.
Maren, S., and Quirk, G.J. (2004). Neuronal signalling of fear memory. Nat. Rev. Neurosci. 5,
844-852.
Margules, D.L., and Olds, J. (1962). Identical "Feeding" and "Rewarding" Systems in the Lateral
Hypothalamus of Rats. Science 135, 374-375.
McKernan, M.G., and Shinnick-Gallagher, P. (1997). Fear conditioning induces a lasting
potentiation of synaptic currents in vitro. Nature 390, 607-611.
Mora, F., Rolls, E.T., and Burton, M.J. (1976). Modulation during learning of the responses of
neurons in the lateral hypothalamus to the sight of food. Exp. Neurol. 53, 508-519.
Muramoto, K., Ono, T., Nishijo, H., and Fukuda, M. (1993). Rat amygdaloid neuron responses
during auditory discrimination. Neuroscience 52, 621-636.
Namburi, P., Beyer, A., Calhoon, G., Yorozu, S., Halbert, S., Felix-Ortiz, A., Wichmann, R.,
Holden, S., Mertens, K., Anahtar, M., et al. (2015). A circuit mechanism for differentiating
positive and negative associations. Nature In press.
52
Nieh, E.H., Matthews, G.A., Allsop, S.A., Presbrey, K.N., Leppla, C.A., Wichmann, R., Neve,
R., Wildes, C.P., and Tye, K.M. (2015). Decoding Neural Circuits that Control Compulsive
Sucrose Seeking. Cell 160, 528-541.
Niu, J.-G., Yokota, S., Tsumori, T., Oka, T., and Yasui, Y. (2012). Projections from the anterior
basomedial and anterior cortical amygdaloid nuclei to melanin-concentrating hormonecontaining neurons in the lateral hypothalamus of the rat. Brain Res. 1479, 31-43.
Olds, J., and Milner, P. (1954). Positive reinforcement produced by electrical stimulation of
septal area and other regions of rat brain. J. Comp. Physiol. Psychol. 47, 419-427.
Pape, H.-C., and Pare, D. (2010). Plastic Synaptic Networks of the Amygdala for the
Acquisition, Expression, and Extinction of Conditioned Fear. Physiol. Rev. 90, 419-463.
Paton, J.J., Belova, M.A., Morrison, S.E., and Salzman, C.D. (2006). The primate amygdala
represents the positive and negative value of visual stimuli during learning. Nature 439, 865-
870.
Petrovich, G.D., and Gallagher, M. (2003). Amygdala Subsystems and Control of Feeding
Behavior by Learned Cues. Ann. N. Y. Acad. Sci. 985, 251-262.
Petrovich, G.D., Canteras, N.S., and Swanson, L.W. (2001). Combinatorial amygdalar inputs to
hippocampal domains and hypothalamic behavior systems. Brain Res. Rev. 38, 247-289.
Petrovich, G.D., Setlow, B., Holland, P.C., and Gallagher, M. (2002). Amygdalo-Hypothalamic
Circuit Allows Learned Cues to Override Satiety and Promote Eating. J. Neurosci. 22, 8748-
8753.
Petrovich, G.D., Holland, P.C., and Gallagher, M. (2005). Amygdalar and Prefrontal Pathways to
the Lateral Hypothalamus Are Activated by a Learned Cue That Stimulates Eating. J. Neurosci.
25, 8295-8302.
Pettit, H.O., and Justice Jr., J.B. (1989). Dopamine in the nucleus accumbens during cocaine
self-administration as studied by in vivo microdialysis. Pharmacol. Biochem. Behav. 34, 899-
904.
Quirk, G.J., Repa, C., and LeDoux, J.E. (1995). Fear conditioning enhances short-latency
auditory responses of lateral amygdala neurons: parallel recordings in the freely behaving rat.
Neuron 15, 1029-1039.
Quirk, G.J., Armony, J.L., and LeDoux, J.E. (1997). Fear conditioning enhances different
temporal components of tone-evoked spike trains in auditory cortex and lateral amygdala.
Neuron 19, 613-624.
Rodin, J. (1981). Current status of the internal-external hypothesis for obesity: what went wrong?
Am. Psychol. 36, 361-372.
53
Roesch, M.R., Calu, D.J., Esber, G.R., and Schoenbaum, G. (2010). Neural correlates of
variations in event processing during learning in basolateral amygdala. J. Neurosci. 30, 2464-
2471.
Rogan, M.T., Staubli, U.V., and LeDoux, J.E. (1997). Fear conditioning induces associative
long-term potentiation in the amygdala. Nature 390, 604-607.
Rumpel, S., LeDoux, J., Zador, A., and Malinow, R. (205). Postsynaptic receptor trafficking
underlying a form of associative learning. Science 308, 83-88.
Sah, P., Faber, E.S.L., Armentia, M.L.D., and Power, J. (2003). The Amygdaloid Complex:
Anatomy and Physiology. Physiol. Rev. 83, 803-834.
Sakurai, T. (2007). The neural circuit of orexin (hypocretin): maintaining sleep and wakefulness.
Nat. Rev. Neurosci. 8, 171-181.
Sakurai, T., Nagata, R., Yamanaka, A., Kawamura, H., Tsujino, N., Miuraki, Y., Kageyama, H.,
Kunita, S., Takahashi, S., Goto, K., et al. (2005). Input of Orexin/Hypocretin Neurons Revealed
by a Genetically Encoded Tracer in Mice. Neuron 46, 297-308.
Sanghera, M.K., Rolls, E.T., and Roper-Hall, A. (1979). Visual responses of neurons in the
dorsolateral amygdala of the alert monkey. Exp. Neurol. 63, 610-626.
Schoenbaum, G., Chiba, A.A., and Gallagher, M. (1998). Orbitofrontal cortex and basolateral
amygdala encode expected outcomes during learning. Nat. Neurosci. 1, 155-159.
Schoenbaum, G., Saddoris, M.P., and Stalnaker, T.A. (2007). Reconciling the Roles of
Orbitofrontal Cortex in Reversal Learning and the Encoding of Outcome Expectancies. Ann. N.
Y. Acad. Sci. 1121, 320-335.
Schultz, W. (2006). Behavioral theories and the neurophysiology of reward. Annu. Rev. Psychol.
57, 87-115.
Schultz, W., Dayan, P., and Montague, P.R. (1997). A neural substrate of prediction and reward.
Science 275, 1593-1599.
Schwartzbaum, J.S. (1988). Electrophysiology of taste, feeding and reward in lateral
hypothalamus of rabbit. Physiol. Behav. 44, 507-526.
Setlow, B., Holland, P.C., and Gallagher, M. (2002). Disconnection of the basolateral amygdala
complex and nucleus accumbens impairs appetitive pavlovian second-order conditioned
responses. Behav. Neurosci. 116, 267-275.
Simon, H., Le Moal, M., and Calas, A. (1979). Efferents and afferents of the ventral tegmentalA10 region studied after local injection of [3H]leucine and horseradish peroxidase. Brain Res.
178, 17-40.
Singh, J., Desiraju, T., and Raju, T.R. (1996). Comparison of Intracranial Self-Stimulation
Evoked From Lateral Hypothalamus and Ventral Tegmentum: Analysis Based on Stimulation
Parameters and Behavioural Response Characteristics. Brain Res. Bull. 41, 399-408.
Smith, A.C., Frank, L.M., Wirth, S., Yanike, M., Hu, D., Kubota, Y., Graybiel, A.M., Suzuki,
W.A., and Brown, E.N. (2004). Dynamic analysis of learning in behavioral experiments. J.
Neurosci. 24, 447-461.
Stuber, G.D., Sparta, D.R., Stamatakis, A.M., van Leeuwen, W.A., Hardjoprajitno, J.E., Cho, S.,
Tye, K.M., Kempadoo, K.A., Zhang, F., Deisseroth, K., et al. (2011). Excitatory transmission
from the amygdala to nucleus accumbens facilitates reward seeking. Nature 475, 377-380.
Ter Horst, G.J., and Luiten, P.G.M. (1987). Phaseolus vulgaris leuco-agglutinin tracing of
intrahypothalamic connections of the lateral, ventromedial, dorsomedial and paraventricular
hypothalamic nuclei in the rat. Brain Res. Bull. 18, 191-203.
Tye, K.M., and Janak, P.H. (2007). Amygdala Neurons Differentially Encode Motivation and
Reinforcement. J. Neurosci. 27, 3937-3945.
Tye, K.M., Stuber, G.D., de Ridder, B., Bonci, A., and Janak, P.H. (2008). Rapid strengthening
of thalamo-amygdala synapses mediates cue-reward learning. Nature 453, 1253-1257.
Tye, K.M., Cone, J.J., Schairer, W.W., and Janak, P.H. (2010). Amygdala Neural Encoding of
the Absence of Reward during Extinction. J. Neurosci. 30, 116-125.
Tye, K.M., Prakash, R., Kim, S.-Y., Fenno, L.E., Grosenick, L., Zarabi, H., Thompson, K.R.,
Gradinaru, V., Ramakrishnan, C., and Deisseroth, K. (2011). Amygdala circuitry mediating
reversible and bidirectional control of anxiety. Nature 471, 358-362.
Uwano, T., Nishijo, H., Ono, T., and Tamura, R. (1995). Neuronal responsiveness to various
sensory stimuli, and associative learning in the rat amygdala. Neuroscience 68, 339-361.
Wagner, A., Aizenstein, H., Venkatraman, V.K., Fudge, J., May, J.C., Mazurkewicz, L., Frank,
G.K., Bailer, U.F., Fischer, L., Nguyen, V., et al. (2007). Altered reward processing in women
recovered from anorexia nervosa. Am. J. Psychiatry 164, 1842-1849.
Wilensky, A.E., Schafe, G.E., and LeDoux, J.E. (1999). Functional Inactivation of the Amygdala
before But Not after Auditory Fear Conditioning Prevents Memory Formation. J. Neurosci. 19,
RC48-RC48.
Wrase, J., Schlagenhauf, F., Kienast, T., Wtistenberg, T., Bermpohl, F., Kahnt, T., Beck, A.,
Str6hle, A., Juckel, G., Knutson, B., et al. (2007). Dysfunction of reward processing correlates
with alcohol craving in detoxified alcoholics. Neurolmage 35, 787-794.
Yizhar, 0., Fenno, L.E., Prigge, M., Schneider, F., Davidson, T.J., O'Shea, D.J., Sohal, V.S.,
Goshen, I., Finkelstein, J., Paz, J.T., et al. (2011). Neocortical excitation/inhibition balance in
information processing and social dysfunction. Nature 477, 171-178.
Zamble, E. (1973). Augmentation of eating following a signal for feeding in rats. Learn. Motiv.
4, 138-147.