Behavioral and Electrophysiological Properties of Nucleus Reuniens: Role in Arousal, Spatial Navigation and Cognitive Processes

BEHAVIORAL AND ELECTROPHYSIOLOGICAL PROPERTIES OF

NUCLEUS REUNIENS: ROLE IN AROUSAL, SPATIAL NAVIGATION AND

COGNITIVE PROCESSES

Tatiana Danela Viena

A Dissertation Submitted to the Faculty of

Charles E. Schmidt College of Science

In Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

Florida Atlantic University

Boca Raton, FL

December 2018

ACKNOWLEDGEMENTS

First, I wish to express my sincere thanks to my advisor Dr. Robert P. Vertes for the opportunity to work on these experiments and for his guidance and patience in completing this project. I also would like to thank the members of my committee, Dr.

Robert W. Stackman and Dr. Timothy A. Allen for their challenging questions and suggestions regarding this project, but more importantly for their support. In addition, I want to thank Dr. Stephanie Linley for patiently teaching, guiding and empowering me whenever it was needed.

I am grateful to all of the undergraduate research assistants who volunteered their time to assist in the many aspects of this project. They are too many to be named here, but you know who you are! I also would like to extend my gratitude to all of the wonderful individuals who one way or another assisted me over my graduate years and the many programs that sponsored my research activities.

I also want to give a special recognition to my family (Bernardo, Randee and

Elysse) for their unconditional love and support; and my mother and her daily words of encouragement. Thanks for always believing in me and pushing me to get to the finish line. Finally, I want to thank God for this journey and for giving me the patience and strength to accomplish this goal.

All of you contributed to my success and for that, I humbly THANK YOU!

ABSTRACT

Author: Tatiana Danela Viena

Title: Behavioral and Electrophysiological Properties of Nucleus Reuniens: Role in Arousal, Spatial Navigation and Cognitive Processes

Institution: Florida Atlantic University

Dissertation Advisor: Dr. Robert P. Vertes

Degree: Doctor of Philosophy

Year: 2018

The hippocampal-medial prefrontal circuit has been shown to serve a critical role in decision making and goal directed actions. While the hippocampus (HF) exerts a direct influence on the medial prefrontal cortex (mPFC), there are no direct return projections from the mPFC to the HF. The nucleus reuniens (RE) of the midline thalamus is strongly reciprocally connected with the HF and mPFC and represents the major link between these structures.

We investigated the role of RE in functions associated with the hippocampus and the mPFC -- or their interactions. Using two different inactivation techniques

(pharmacological and chemogenetic), we sought to further define the role of RE in spatial working memory (SWM) and behavioral flexibility using a modified delayed non-match to sample (DNMS) working memory task. We found that the reversible inactivation of

RE with muscimol critically impaired SWM performance, abolished well-established spatial strategies and produced a profound inability to correct non-rewarded, incorrect

choices on the T-maze (perseverative responding). We observed similar impairments in

SWM following the chemogenetic (DREADDs) inactivation of RE or selective RE projections to the ventral HF. In addition, we showed that the inhibition of RE terminals to the dorsal or ventral HF altered task related behaviors by increasing or decreasing the time to initiate the task or reach the reward, respectively. Finally, we examined discharge properties of RE cells across sleep-wake states in behaving rats. We found that the majority of RE cells discharge at high rates of activity in waking and REM and at significantly reduced rates in SWS, with a subpopulation firing rhythmically in bursts during SWS. We identified five distinct subtypes of RE cells that discharged differently across vigilant states; those firing at highest rates in waking (W1, W2), in REM sleep

(R1, R2) and SWS (S1). Given the differential patterns of activity of these cells, we proposed they may serve distinct functions in waking – and possibly in SWS/REM sleep.

In sum, our findings indicate that RE is critically involved in mnemonic and executive functions and the heterogeneous activity of these cells support a role for RE in arousal/attention, spatial working memory and cognition.

DEDICATION

This manuscript is dedicated to my daughters Randee and Elysse, who I constantly tell they can achieve anything they want in life through honesty, persistence and hard work. Let this be proof and a constant reminder that the sky is the limit, and when you get there, you can still fly!! Therefore, chase your dreams and never, EVER give up!

BEHAVIORAL AND ELECTROPHYSIOLOGICAL PROPERTIES OF

NUCLEUS REUNIENS: ROLE IN AROUSAL, SPATIAL NAVIGATION AND

COGNITIVE PROCESSE

LIST OF TABLES ...... xiv

LIST OF FIGURES ...... xv

ABBREVIATIONS ...... xvii

GENERAL INTRODUCTION ...... 1

The Ventral Midline Thalamus ...... 1

Reuniens and rhomboid nuclei: Inputs/outputs ...... 2

Nucleus reuniens inputs ...... 2

Nucleus reuniens outputs ...... 5

RE connections with mPFC ...... 8

RE connections with HF ...... 8

Rhomboid nucleus...... 9

Arousal related inputs to RE/RH ...... 12

Ventral midline thalamus: Electrophysiological studies ...... 13

Role of nucleus reuniens in mnemonic functions ...... 18

Role of nucleus reuniens in executive functions ...... 20

viii

Role of the thalamus in arousal/attention ...... 24

Purpose and Dissertation Organization ...... 26

CHAPTER 1: INACTIVATION OF NUCLEUS REUNIENS IMPAIRS

SPATIAL WORKING MEMORY AND BEHAVIORAL FLEXIBILITY IN

THE RAT ...... 28

Executive Summary ...... 28

CHAPTER 2: CHEMOGENETIC INACTIVATION OF NUCLEUS

REUNIENS OR RE FIBERS SELECTIVELY DISTRIBUTING TO THE

MPFC AND HIPPOCAMPUS: MEMORY AND GOAL DIRECTED

ACTIONS ...... 30

Introduction ...... 30

The role of hippocampal and prefrontal interactions in the delayed nonmatch

to sample paradigm...... 31

RE as a mediator of HF-mPFC interactions in cognition ...... 33

Materials and Methods ...... 34

Subjects ...... 34

DREADDs Technique ...... 34

Surgery and Viral Vector Delivery ...... 35

Drugs and microinfusion procedures...... 36

Apparatus ...... 39

Behavioral Testing ...... 39

Data Analysis...... 43

Coding of behavioral data ...... 43

Performance measures...... 43

Immunohistology ...... 44

Statistical Analysis ...... 46

Results ...... 47

DREADDs viral expression and histological verification ...... 47

Chemogenetic inhibition of RE alters choice accuracy across delays but does

not produce spatial perseveration ...... 55

Chemogenetic inactivation of RE terminal projections to the vHF impairs

working memory performance at intermediate delays...... 58

Chemogenetic inhibition of the RE > dHF pathway does not alter choice

accuracy, or produce perseveration...... 58

Chemogenetic inhibition of the RE > mPFC pathway does not impair choice

accuracy nor produce perseveration ...... 59

Systemic CNO inactivation and CNO inhibition of RE terminals in the

mPFC, dHF and vHF did not produce win-shift errors between trials ...... 64

Inactivation of RE terminals in the dHF or vHF, but not in the mPFC, alters

latencies to initiate actions and time to reach reward site on the DNMS task ...... 64

T-maze start box (SB)...... 64

Sample runs ...... 64

Choice runs ...... 65

T-maze junction ...... 65

T-maze goal area ...... 68

Sample runs ...... 68

Choice runs ...... 68

Locomotive speed ...... 69

Discussion...... 72

RE > ventral hippocampal pathway: critical for SWM and goal directed

behaviors...... 73

Chemogenetic inactivation of RE or RE projections to the hippocampus or

prefrontal cortex did not produce spatial perseveration in the DNMS task ...... 76

Conclusions ...... 77

CHAPTER 3: CHARACTERIZATION OF UNITS OF NUCLEUS REUNIENS

ACROSS SLEEP AND WAKING STATES ...... 79

Introduction ...... 79

Role of the thalamus in the modulation of the cortical EEG during wake-

sleep states ...... 79

Activity of thalamic nuclei in sleep and arousal ...... 79

Role of nucleus reuniens in sleep-waking and attention...... 81

Materials and Methods ...... 82

Subjects ...... 82

Surgical Procedures ...... 83

Electrode construction ...... 84

Apparatus ...... 87

Procedures ...... 87

Sleep Deprivation (SD) ...... 88

Electrophysiological recordings ...... 88

Single Units ...... 89

Cortical and hippocampal EEG...... 89

Histology ...... 89

Data Analysis...... 90

Sleep Stage Coding Criteria ...... 90

Single unit waveform sorting ...... 91

Spectral analysis of the cortical and hippocampal EEG ...... 91

Separation of unit spikes across vigilance states ...... 91

Firing discharge patterns ...... 94

Spike waveforms...... 94

Separation by discharge profile across states ...... 94

Statistical analysis ...... 94

Results ...... 95

Histological verification ...... 95

RE discharge patterns and waveforms ...... 95

Comparison of discharge profiles of RE neurons across wake-sleep states—

with wake-active (W) and REM-active (R) cells subdivided into W1, W2, R1

and R2 type cells ...... 107

Patterns of RE activity during active W, NREM and REM sleep ...... 110

Discussion...... 115

xii

Diversity of RE activity across vigilant states ...... 115

Conclusions ...... 118

GENERAL DISCUSSION ...... 121

RE as a conduit in the exchange of cognitive information between the

hippocampus and the medial prefrontal cortex in spatial working memory ...... 122

Summary ...... 130

APPENDICES ...... 133

Appendix A ...... 134

Appendix B ...... 149

REFERENCES ...... 150

xiii

LIST OF TABLES

Table 1. Mean latencies at critical points in the T-maze during performance of the

DNMS task following CNO microinjections in RE target sites...... 66

Table 2. Mean speeds during performance on the DNMS T-maze task following

CNO injections in RE target sites...... 70

Table 3. Mean discharge rates of RE cells units across sleep-wake states ...... 98

Table 4. Mean discharge rates of RE cells classified by waveforms across sleep-wake

states ...... 105

Table 5. Mean discharge for RE cells classified according to discharge profile across

sleep-wake states...... 108

xiv

LIST OF FIGURES

Figure 1. Neuroanatomical organization of the midline thalamus...... 3

Figure 2. Major inputs and outputs of nucleus reuniens...... 6

Figure 3. Nucleus reuniens-medial prefrontal–hippocampal (PFC-HF) circuit...... 10

Figure 4. Evoked responses in the mPFC elicited by stimulation of RE...... 16

Figure 5. RE involvement in strategy shifting using a double-H water maze...... 22

Figure 6. Set up for chemogenetically (with DREADDs) inactivating RE and its

main targets...... 37

Figure 7. Schematic representation of procedures for testing the effects of DREADD

injections in RE using the delayed non-match to sample (DNMS) T-maze task...... 41

Figure 8. Immunoreacted sections through the thalamus showing DREADD-infected

cells in RE...... 48

Figure 9. Distribution of mCherry-immunolabeled RE fibers in target structures...... 51

Figure 10. Nissl stained and associated schematic coronal sections showing cannula

placements in the mPFC, dHF and vHF...... 53

Figure 11. Effects of DREADD suppression of RE on choice behavior in DNMS T-

maze task at two delay times...... 56

Figure 12. Effects of DREADD suppression of RE terminals on choice behavior in

the DNMS T-maze task...... 60

Figure 13. Effects of DREADD suppression of RE as well as RE terminals to targets

on perseverative behavior...... 62

Figure 14. Components of customized drivable microelectrode assembly...... 85

Figure 15. Method for sleep deprivation and behavioral/EEG signs of wake-sleep

states...... 92

Figure 16. Sites of unit recording in RE and slow wave recordings in the dorsal

hippocampus...... 96

Figure 17. Mean firing rates of RE cells across waking, SWS and REM sleep...... 101

Figure 18. Types of spike waveforms of RE units...... 103

Figure 19. Discharge patterns of two types of RE cells across W, SWS and REM

sleep...... 111

Figure 20. Discharge patterns of three types of RE cells across W, SWS and REM

sleep...... 113

xvi

ABBREVIATIONS

5-HT Serotonergic

AAV Adeno-associated virus

ARAS Ascending reticular activating system

CA1 Field CA1, Ammon’s horn

CA2 Field CA2, Ammon’s horn

CA3 Field CA3, Ammon’s horn

CED Cambridge Electronics Design

CLA Claustrum

CM Central medial

CNO Clozapine-n-oxide dCA1 Dorsal CA1 dHF Dorsal hippocampus

DMTS Delayed match to sample

DNMTP Delayed non-match to position

DNMS Delayed non match to sample

DR Dorsal raphe nucleus

DREADD Designer receptors exclusively activated by designer drugs

EEG Electroencephalogram fEPSP Field excitatory postsynaptic potentials

GA Goal Area

xvii

Gi G coupled receptor i

HF Hippocampus hM4Di Human M4 muscarinic receptor

I.P Intraperitoneal

IL Infralimbic cortex

IMD Interomediodorsal nucleus

INT Intersection

ITI Intertrial interval

LC Locus coeruleus

LDT Laterodorsal tegmentum

LFP Local field potentials

LGN Lateral geniculate nuclei

LTP Long-term potentiation

MO Medial orbital cortex mPFC Medial prefrontal cortex

MR Medial raphe nucleus

MT Medial or midline thalamus

NAc Nucleus accumbens

NREM Non-REM sleep

OC Orbital cortex

PFC Prefrontal cortex

PL Prelimbic cortex

PP Paired pulse

xviii

PPT Pedunculopontine tegmental

RAM Radial arm maze

RE Nucleus reuniens of thalamus

REM Rapid-eye-movement, paradoxical sleep

RH Rhomboid nucleus of thalamus

RT Reticular thalamic nucleus

S-W Sleep-wake

S.C Subcutaneous

SB Startbox

SD Sleep deprivation

SLM Stratum lacunosum moleculare

SS Stainless steel

Sub Subiculum

SWM Spatial working memory

SWS Slow-Wave Sleep

TC Thalamocortical cells, thalamic cells

TD Trajectory-dependent vCA1 Ventral CA1 vHF Ventral hippocampus

VMT Ventral midline Thalamus

W Active Wake

WM Working memory

xix

GENERAL INTRODUCTION

The Ventral Midline Thalamus

The midline thalamus was originally considered part of the ‘non-specific’ thalamus due to its widespread and less specific connections with the cortex. This dramatically contrasted with first order thalamic (relay) nuclei that generally target

“specific” sensorimotor cortical sites. More recently, as the connections and functions of the midline thalamus have become more fully understood, alternative classifications for the thalamus have been proposed which no longer view the medial/midline thalamus as

“non-specific”.

A recent classification divides the thalamus into three distinct groups: (1) sensorimotor nuclei, which include the traditional first-order relay nuclei; (2) a “limbic” group consisting of the midline, intralaminar and anterior nuclei; and (3) ‘bridging’ nuclei that combine the sensorimotor and limbic domains – mainly thalamic nuclei located in lateral portions of the thalamus (Vertes et al., 2015a, 2015b).

The ‘limbic thalamus’, then, is composed of the anterior, intralaminar and midline nuclei and additionally the mediodorsal nucleus (MD) of the thalamus. The midline nuclei are further divided into a dorsal midline group, which includes the paratenial (PT) and paraventricular (PV) nuclei, and a ventral midline group (VMT) composed of the nucleus reuniens (RE) and rhomboid (RH) nucleus of thalamus. The VMT is located along the most ventral and medial portion of the thalamus. At the rostral thalamus, the

VMT is located ventral to PT and PV. At the mid-thalamus, RE/RH lie ventral to PV,

MD, the interanteromedial (IAM) and central medial (CM) nuclei of thalamus. Caudally, the VMT is positioned beneath the interomediodorsal nucleus (IMD) and CM. A notable anatomical landmark of VMT is that along its rostro-caudal extent, it lies dorsal to the third ventricle (Vertes, 2002; Cassel et al., 2013). The anatomical location of RE/RH within the midline thalamus is illustrated in Figure 1 (adapted from Cassel et al., 2013).

Reuniens and rhomboid nuclei: Inputs/outputs

The VMT receives a vast set of afferents from the brainstem, the diencephalon and telencephalon, most originating from limbic or limbic-associated sites (McKenna and

Vertes, 2004; Vertes, 2002; 2004; Vertes et al., 2015a, 2015b).

Nucleus reuniens inputs

The brainstem is a rich source of input to RE. RE receives afferents from the laterodorsal (LDT) and pedunculopontine tegmental (PPT) nuclei, locus coeruleus (LC), dorsal raphe nuclei (DR), substantia nigra (SN; pars compacta), parabrachial nuclei (PB) and the ventral tegmental area (VTA; McKenna and Vertes, 2004; Vertes et al., 2010).

Interestingly, most of these nuclei are part of the ascending reticular activating system

(ARAS) which is positioned to modulate levels of arousal in sleep and waking.

The diencephalic afferents to RE mainly arise from the lateral habenula, the subthalamus, and from several groups of the hypothalamus including the anterior, dorsomedial, ventromedial, lateral mammillary, supramammillary (SUM) and tuberomammillary nuclei (McKenna and Vertes, 2004). RE also receives afferents from the reticular nucleus (RT) of thalamus and the zona incerta (ZI). As RE neurons are almost exclusively excitatory, RT and ZI represent major sources of inhibitory input to reuniens (Herkerman, 1978; Vertes et al., 2006). Other subcortical sites providing quite

Figure 1 Neuroanatomical organization of the midline thalamus. A) Ventral midline thalamic structures, nucleus reuniens (pink) (Re) and the rhomboid nucleus (green) (Rh) at three anterior-posterior levels of the thalamus B) RE connections with its main cortical targets, the medial prefrontal cortex (mPFC) and hippocampus (HIP). Abbreviations: CEM, central medial nucleus of thalamus; IAM, interanteromedial nucleus of thalamus; IMD, interomediodorsal nucleus of thalamus; MD, mediodorsal nucleus of thalamus; pRe, perireuniens nucleus; PT, paratenial nucleus of thalamus; PV, paraventricular nucleus of thalamus; Re, nucleus reuniens; Rh, rhomboid nucleus. Adapted from Cassel et al., 2013; with permission.

Figure 1

prominent input to RE include the amygdala, claustrum, and the lateral regions of the basal forebrain (Canteras and Swanson, 1992; McKenna and Vertes, 2004; Witter et al., 1990).

Hippocampal input to RE is pronounced and mainly originates from the dorsal and ventral subiculum and to a much lesser extent from ventral CA1 and the pre- and post-subiculum (Herkenham, 1978; Cassel et al., 2013). A very prominent source of cortical afferents to RE is the medial prefrontal cortex (mPFC), and all divisions of the mPFC: the medial agranular (AGm), anterior cingulate (AC), prelimbic (PL) and infralimbic (IL) cortices. RE also receives moderate to heavy input from the medial orbital, retrosplenial, perirhinal and entorhinal (EC) cortices (McKenna and Vertes, 2004;

Vertes et al., 2007). Figure 2 depicts RE major inputs and outputs.

Nucleus reuniens outputs

Whereas RE receives a diverse and vast set of afferent projections (see above), the output of RE is relatively restricted, mainly targeting cortical as opposed to subcortical sites and limbic cortical structures (Vertes et al., 2015a, 2015b). Subcortically, RE distributes moderately to the nucleus accumbens (NAc) and claustrum, and lightly to regions of the amygdala and hypothalamus (Herkenham, 1978; Ohtake and Yamada,

1989; Vertes, 2004; Vertes et al., 2006, 2015a, 2015b; Cassel et al., 2013).

Whereas, as discussed, the input to RE from HF mainly arises from the dorsal and ventral subiculum, RE return projections to the hippocampus distribute more widely, strongly innervating the dorsal and ventral CA1, the dorsal and ventral subiculum and the pre- and parasubiculum (McKenna and Vertes, 2004; Vertes, 2006, Hoover and Vertes,

2012; Cassel et al., 2013). RE distributes to virtually all regions of “limbic cortices”.

Rostrally, this would include the medial (MO) ventral (VO) and ventrolateral (VLO)

Figure 2 Major inputs and outputs of nucleus reuniens. Top: Major inputs to RE from various structures of the brainstem, the diencephalon, and the telencephalon, prominently including those from the cortex. Bottom: By comparison, the output is more limited mainly distributing to the hippocampus and the prefrontal-orbital cortex. Abbreviations: LH, lateral habenula; HYP, hypothalamus; SUM, supramammillary bodies; ZI, zona incerta; LD, lateral dorsal nucleus of thalamus; LGN, lateral geniculate nucleus of thalamus; PV, paraventricular of thalamus; TRN, thalamic reticular nucleus; RF, reticular formation; DLT, dorsolateral tegmental nucleus; LC, locus coeruleus; PN, parabrachial nucleus; RN, raphe nuclei; PPT, pedunculopontine tegmental nucleus; VTA, ventral tegmental area; SN, suprachiasmatic nucleus; PFC, prefrontal cortex; mAG, medial agranular cortex; ACC, anterior cingulate cortex; IL infralimbic cortex; PL, prelimbic cortex; HC, hippocampus; EC, enthorhinal cortex; PER, perirhinal cortex; Sub, subiculum; AMY, amygdala; BF basal forebrain; CLA, claustrum; MO, medial orbital cortex; VO, ventral orbital cortex; VLO ventrolateral orbital cortex; FPC, frontal polar cortex; PARA, parahippocampal cortex; PIRI, piriform cortex; NAc, nucleus accumbens; dCA1, dorsal CA1; vCA1, ventral CA1. .

Figure 2

orbital cortices, the piriform cortex, and all divisions of the mPFC (AGm, AC, PL, IL) and further caudally mainly parahippocampal cortices: ectorhinal, perirhinal and medial and lateral entorhinal cortices (Herkenham, 1978; Ohtake and Yamada, 1989; Van der

Werf et al., 2002; Vertes et al., 2006, 2015a, 2015b; Hoover and Vertes, 2011, 2012;

Varela et al., 2014).

RE connections with mPFC

In rats, the mPFC, particularly IL and PL, receive substantial input from CA1 and the subiculum of HF (Jay and Witter, 1991; Vertes, 2006; Hoover and Vertes, 2007), but interestingly there are no direct return projections from the mPFC to the HF. Based on the pronounced mPFC projections to RE (see above) and in turn equally strong RE projections to the HF, RE is ideally positioned to relay information from the mPFC to the

HF, thus completing a loop between these structures: HF > mPFC > RE > HF (Vertes,

2002; Vertes et al., 2006, 2007, 2015a, 2015b; Varela et al ., 2014; Griffin, 2015).

RE connections with HF

RE is a major source of thalamic projections to the HF (Su and Bentivoglio, 1990;

Vertes et al., 2015a, 2015b). RE fibers distribute selectively to the stratum lacunosum moleculare (slm) of the dorsal and ventral hippocampus and to the outer moleculare layer of the ventral subiculum. There are no RE projections to CA2, CA3, or to the dentate gyrus (Vertes et al., 2006, 2007, 2015a, 2015b; Cassel et al., 2013; see Figure 3). Vertes et al. (2007) demonstrated that mPFC fibers distributing to RE make excitatory synaptic connections with RE neurons projecting to the HF. It has further recently been shown that approximately 8-10% of RE neurons project, via axon collaterals to the HF and mPFC (Hoover and Vertes, 2012; Varela et al., 2014). RE cells with branches to HF and

the mPFC are found throughout RE but concentrated in mid and lateral portions of RE

(Hoover and Vertes, 2012). This suggests that RE may simultaneously influence HF and the mPFC to coordinate their activity for working memory or other functions (see below).

Finally, it has been demonstrated that RE projections to the ventral HF are ten-fold greater than those to the dorsal hippocampus (Hoover and Vertes, 2012).

Rhomboid nucleus

The rhomboid nucleus (RH) is characterized by its unique rhomboid shape and

“large darkly stained cells” (Cassel et al., 2013). In contrast to RE, the RH has been examined in considerably less detail, possibly owing to the difficulty of analyzing such a relatively small and unusually shaped structure. Similar to RE, RH receives pronounced input from the brainstem primarily from arousal-related structures such as the reticular formation, LDT, PPT, SNpc, LC and the parabrachial complex (Rassnick et al., 1998;

Krout and Loewy, 2000; Vertes et al., 2010; Cassel et al., 2013). Cortical afferents to RH arise from the mPFC (AGm, AC, IL, PL) but to a lesser degree than those to RE.

However, unlike RE, RH also receives projections from the primary and secondary motor cortices, and the primary somatosensory cortex (Vertes, 2002, Vertes, 2004). No hippocampal projections to RH have been identified (Cassel et al., 2013).

RH targets most of the same cortical areas from which it receives projections and is thus reciprocally connected with the orbitomedial PFC and sensory-motor cortices

(McKenna and Vertes, 2004). Whereas RH does not receive input from HF, RH projects to the hippocampus -- in generally the same pattern, but to a much lesser degree than RE

(Vertes et al., 2006). The demonstration that RH, like RE, links the mPFC with the HF suggests that RH may complement the actions of RE within this circuit, and

Figure 3 Nucleus reuniens-medial prefrontal–hippocampal (PFC-HF) circuit. The ventral hippocampus (vCA1) sends direct projections to the infralimbic and prelimbic cortices (IL, PL; in green) of the mPFC. There are no direct projections from the mPFC to the HF. In lieu of that, RE is the main relay from the mPFC to the HF. RE is strongly reciprocally connected with the mPFC and HF. Arrows depicted in purple: RE projections to IL/PL, dorsal (dCA1) and vCA1. Red arrows: IL/PL projections to RE. Blue arrows: HF projections to RE. Red and blue dots represent RE terminal projections to the mPFC and HF, respectively.

Figure 3

additionally, RH may provide important sensori-motor information to the HF in the regulation of various aspects of behavior.

Arousal related inputs to RE/RH

Afferents to RE/RH from the brainstem suggests that various monoaminergic

(serotonergic, noradrenergic, dopaminergic), as well as cholinergic systems are important for the regulation of the RE/RH activity (Krout et al., 2002; McKenna and Vertes, 2004).

For instance, it is known that ascending 5-HT fibers from the dorsal (DR) and median raphe (MR) nuclei distribute heavily to RE/RH (Moore et al., 1978; Vertes, 1991; Vertes et al., 1999, 2010), 5-HT7 receptors are highly expressed in midline thalamic nuclei

(Gustafson et al., 1996; Vizuete et al., 1997), and stimulation of cholinergic LDT and

PPT neurons modulates midline thalamic activity (Pare and Steriade, 1990; Steriade,

1996). It is well known that monoaminergic and cholinergic systems are directly involved in the regulation of sleep-waking states -- as well as several other functions including attention, working memory and affective behaviors. Relevant information from these brainstem systems is possibly integrated into the mPFC-HF network, via RE/RH, to assist in the execution of affective and cognitive functions requiring increased levels of attention.

The foregoing, then, indicates that RE/RH is a pivotal link to the hippocampus; not only relaying information from the mPFC to the HF, but also from arousal-related brainstem and diencephalic sites that, in part, may modulate the transfer of signals from mPFC to HF, via RE. As RE/RH are positioned to significantly affect emotional and cognitive behaviors, further exploring the role of RE in arousal, memory, and executive functions could provide a better understanding of the influence of RE on circuits essential

for decision-making and goal directed behaviors.

Ventral midline thalamus: Electrophysiological studies

Although numerous studies have described the anatomy of VMT, the electrophysiological properties of RE neurons have remained largely unexplored. Until recently, only a few reports have examined the activity of RE cells.

Pioneering work of Dollerman-Van Weel et al (1997) demonstrated that electrical stimulation of RE had a dual influence at the dorsal CA1. Specifically, using in vivo anesthetized animals, RE stimulation was shown to evoke a local depolarization (sink) at apical dendrites of CA1 pyramidal cells, as well as, an excitation of non-pyramidal cells in strata oriens/alveus of CA1. Of note, spikes were only generated in CA1 with high intensity, low frequency (0.3-2 Hz) RE stimulation, but not with stimulation at theta frequencies (5-10 Hz). It was further shown that paired pulse (PP) stimulation of RE, at low and high intensities, produced enhanced field excitatory postsynaptic potentials

(fEPSP) in the pyramidal cell layer of CA1. Finally, rostral RE stimulation evoked fEPSPs at CA1 at monosynaptic latencies, while caudal RE stimulation elicited them at disynaptic latencies – suggesting that caudal RE effects are channeled through the rostral

RE (Dollerman-Van Weel et al., 1997).

Shortly thereafter, Bertram and Zhang (1999), using a comparable in vivo model, showed that PP stimulation of RE produced marked paired pulse facilitation (PPF) at

CA1, which was in fact, more pronounced than that elicited at CA1 with CA3 stimulation. They further importantly demonstrated that high frequency stimulation of RE generated long-term potentiation (LTP) at CA1, which was sustained for relatively long periods of time. These findings indicate that RE exerts a powerful excitatory effect on

the HF -- at least equivalent to (or possibly more so) than that produced by an intra- hippocampal circuitry (CA3 to CA1; Bertram and Zhang, 1999; Hoover and Vertes,

2012).

Witter and colleagues (Dollerman-Van Weel et al., 2017) recently examined the influence of RE on entorhinal cortical (EC)-elicited evoked responses at the slm layer of

CA1. Using an acute in vivo model, they demonstrated that low frequency (0.3-2 Hz), simultaneous (or joint) stimulation of RE and EC elicited a much greater enhancement of evoked potentials at CA1 (non-linear effects) than those produced by separately stimulating RE and EC – or the sum of the two effects. These findings suggest that RE may be instrumental in modulating (or gating) the flow of information from the EC to

CA1 as the combined activation of RE and EC had a much greater impact at CA1 than

EC stimulation alone.

Only a few studies have examined the electrophysiological characteristics of RE cells (units) in freely moving rats. Jankowski and colleagues (Jankowski et al., 2014,

2015) described populations of RE cells with spatial properties -- similar to “spatial cells” of the HF and other structures. Specifically, they initially (Jankowski et al. 2014) identified head direction (HD) cells in RE and subsequently (Jankowski et al., 2015), place and border cells units in RE – with the latter mainly located dorsally in RE.

Importantly, however, only about 10% of RE neurons exhibited spatial properties, with greater than 64% of cells showing no temporal or spatial characteristics or relationship to the hippocampal EEG/theta (Jankowski et al., 2014, 2015).

Ito et al. (2015) identified a relatively large percentage of RE cells in behaving rats that displayed “trajectory-dependent” (TD) firing properties – similar to those

initially described in the hippocampus (Frank et al., 2000; Wood et al., 2000). These cells “signal” future directional heading as they fire differentially in the central stem of T- maze depending on left or right turns on the maze – preceding and hence “predicting” choice behavior (Ito et al. 2015). Consistent with the finding that RE targets CA1 but not

CA3 (Vertes et al., 2006), they observed a considerably higher percentage of TD cells in

CA1 than CA3, and further showed that the trajectory-dependent firing of CA1 neurons was significantly suppressed with lesioning or optogenetic silencing of RE cells (Ito et al.

2015). It was suggested that TD information (prospective choice) is conveyed from the mPFC to the HF via RE.

To our knowledge, only a single report has examined the electrophysiological characteristics of RE neurons in an in vitro slice preparation (Walsh et al., 2017). Using this technique, the authors demonstrated that RE neurons: (1) discharge spontaneously at rest at about 8 Hz; (2) show a range of firing profiles; (3) have high input resistance; and

(4) display an intrinsic theta rhythmicity. In addition, a small population of RE neurons fired in short, high frequency bursts.

Whereas the mPFC, together with the HF, is a major RE target, few studies have examined the physiological effects of RE on the mPFC. Viana Di Prisco and Vertes

(2006) showed that RE stimulation produced monosynaptic evoked potentials dorsoventrally throughout the mPFC, with most pronounced effects (short latency, large amplitude) at the ventral mPFC (IL/PL). RE stimulation also resulted in strong PP facilitation at IL/PL (see Figure 4).

In summary, the foregoing indicates that RE exerts strong excitatory actions on the hippocampus and the mPFC and although limited, recent studies indicate that RE

Figure 4 Evoked responses in the mPFC elicited by stimulation of RE. Evoked responses at three rostral to caudal levels (A-C) of the medial prefrontal cortex following stimulation of RE. As demonstrated, highest amplitude monosynaptic evoked potentials were present ventrally in the mPFC at the infralimbic and prelimbic cortex. Abbreviations: AC, anterior cingulate cortex; AGl, lateral agranular cortex; AGm, medial agranular cortex; AIv, ventral agranular insular cortex; C-P, caudate-putamen; IL, infralimbic cortex; PIR, piriform cortex; PL, prelimbic cortex; OT, olfactory tubercle; TTd, tenia tecta, dorsal part. Reproduced from Viana Di Prisco and Vertes, 2006, with permission.

Figure 4

cells exhibit various spatial properties -- suggestive of a direct RE influence on the hippocampus in spatial navigation and spatial learning.

Role of nucleus reuniens in mnemonic functions

In the recent past, several studies have assessed the role of RE in mnemonic functions. Early work examined the involvement of RE in the acquisition, retention and retrieval of memories (Davoodi et al., 2009; Dolleman-van der Weel et al. 2009;

Hembrook and Mair, 2011). Based on the strong reciprocal connections of RE with the

HF and mPFC, Hembrook et al. (2012) examined the influence of RE on tasks involving the coordinated actions of the HF and the mPFC. Specifically, using a delayed non- match to position (DNMTP) task, which engages the HF and the mPFC, and a varying choice radial maze (RAM) delayed nonmatching (VC-DNM) task that only recruits the

HF, they demonstrated that lesions of RE only affected performance in the DNMTP task, but not the visuospatial one. The authors concluded that RE was essential to spatial working memory (SWM), but only in tasks that engaged the HF and the mPFC

(Hembrook et al., 2012).

Subsequent studies provided additional support for this idea (Loureiro et al., 2012,

Hallock et al., 2013; Layfield et al., 2015). Notably, Hallock et al. (2013), using a match to position task on a continuous T-maze, examined the role of RE/RH in versions of the task that were dependent on, or independent of, working memory (WM). The WM- dependent task involved the recruitment of the HF and the mPFC, while the non-WM task relied on tactile and visual cues for successful performance. Inactivation of RE/RH with muscimol produced significant deficits on the WM dependent task, without altering performance on the non-WM version of the task (Hallock et al., 2013).

Griffin and colleagues (Layfield et al., 2015) subsequently analyzed RE on a delay dependent alternation task on the T-maze using no delays, short delays (5 s) or intermediate delays (30 s). They found that inactivation of RE significantly altered WM performance at 5 second delays (at the highest dose) and at 30 second delays (at low and high doses), but not for the continuous (no delay) condition. The findings provided further support for the involvement of RE in SWM, and additionally showed that performance progressively deteriorates with increasingly longer delays.

In this regard, we recently showed that the inactivation of RE with muscimol produced significant deficits in a delayed alternation T-maze task at intermediate to long delays (30s, 60s or 120s), and also resulted in pronounced perseverative responses wherein rats repeatedly made incorrect choices on the T-maze during ‘no delay/no- reward’ correction trials (see Viena et al., 2018/Chapter 1 of this dissertation).

In a seminal study, Griffin and colleagues (Hallock et al., 2016) examined the neural interactions of the HF and the mPFC, and the influence of RE, on a delayed alternation T-maze task (see above). They demonstrated that during successful trials on the WM-dependent task: (1) specific populations of mPFC cells were entrained to the hippocampal theta rhythm during the delay; (2) mPFC units fired at highest rates during the holding (delay) period in the start box; (3) mPFC-HF phase coherence in the theta band (but not in delta, beta or gamma bands) was significant at the choice point; and

(4) HF leads mPFC theta during the delays in the start box, while mPFC leads HF slow gamma at the choice point. Importantly, muscimol injections in RE/RH disrupted unit and slow wave interactions of the HF-mPFC, as well as, choice behavior on the delayed alternation T-maze task (Hallock et al., 2016). In sum, the foregoing demonstrates

that RE, as a vital link between the HF and the mPFC, serves a critical role in spatial

WM.

Role of nucleus reuniens in executive functions

In addition to its involvement in SWM, recent evidence indicates that RE also serves a key role in executive functions, such as goal directed behavior and behavioral flexibility. In this regard, an early report by Witter and colleagues (Dolleman Van der

Weel et al., 2009) showed RE lesions spared spatial reference memory on a water maze task, but resulted in a maladaptive search strategy on the probe test of the task. In effect on the probe test, rats immediately swam to the correct quadrant but quickly abandoned this strategy in favor of searching the entire pool for the hidden platform. This was viewed as a prefrontal (PFC), but not hippocampal dysfunction – or an “executive” dysfunction. Supporting this, Chudasama and co-workers (Prasad et al., 2013) demonstrated that RE lesioned rats made premature responses on a 5-choice serial reaction time task (5-CSRTT), indicating a lack of behavioral inhibition which is associated with disruptions of the PFC.

Cholvin et al. (2013) provided strong evidence for a role for RE in adapting to changes in response strategies using a double-H water maze. In this paradigm, the escape platform was always located in the same spatial location (NE corner/arm of the maze; see

Figure 5) and rats were trained to navigate to that platform of the maze. During probe trials, however, rats were released from a different location, such that using a place strategy would lead to successful performance, but a response strategy (a right and then left turn, as previously learned) would lead to the incorrect arm. While control rats were able to shift from a response to place strategy, rats injected with muscimol in RE, or in

the HF and mPFC, continued to use the previously ‘learned’, and now incorrect, response strategy. Cassel and colleagues concluded that the combined actions of the mPFC, (for set-shifting and strategy selection), and the HF (for memory for location) are necessary for spatially guided goal directed behavior, and that RE mediates the exchange of information between these two structures (Cholvin et al., 2013).

Linley et al. (2016) explored the role of RE in executive functions using an attentional set shifting task (AST). In this task, rats are trained to discriminate between odor and tactile cues over several stages of the task which tests for attentional set, attentional set shifting and reversal learning – measures of executive function involving the PFC. Specifically, rats were trained to dig for food rewards in two food cups

(ramekins) placed in separate compartments of the apparatus -- with each ramekin containing different odors and different digging materials (mediums). In a “compound discrimination” (CD) phase of the task, rats were required to select one of the two odors/ramekins that was randomly paired with either of the two mediums. In the CD reversal phase of the task, the previously unrewarded odor was now rewarded, requiring a shift in stimulus-response contingency to a novel odor/tactile combination. Lesions of

RE/RH did not impair simple or compound discrimination, but resulted in significant deficits in reversal learning. In addition, RE lesioned rats showed impairments in the intradimensional shift phase of the task, wherein new combinations of odors and mediums were introduced, but rats were still required to attend to specific odors for reward. The deficits were interpreted as an impairment in rule abstraction, or a failure to use a previously learned rule or strategy to guide behavior (Linley et al., 2016). This dysfunction reflected a loss of behavioral flexibility, which is commonly associated with

Figure 5 RE involvement in strategy shifting using a double-H water maze. Top: Responses learned during training. When released from the north (N) position of the maze, rats learned to make two consecutive left turns to reach the submerged platform at the northeast (NE) corner. When released from the south position (S), rats learned to make a left and right turn to reach the platform. Bottom: During probe trials, rats were released from the southwest (SW) arm. With this condition, if rats used the previously learned right-left sequence they would choose the north (N) arm (response strategy) rather than correctly shifting the strategy to reach the NE arm (place response). Animals that received muscimol injections in RE/RH, the HF or the mPFC, compared to controls, were not able to shift from a response to a place strategy. Adapted from Cassel et al., 2013; with permission.

Figure 5

alterations of the orbitomedial prefrontal cortex (McAlonan and Brown, 2003; Clark et al., 2004, 2005, 2006; Dalley et al., 2004).

Together these studies provide important evidence for a role for RE in executive cognitive control. RE is strongly and reciprocally connected with the mPFC and the orbital cortex and through these connections may influence goal directed behavior by updating context specific information, learned response rules/strategies, and salient reward expectancies/contingencies necessary to successfully perform working memory related tasks.

Role of the thalamus in arousal/attention

While an increasing number of studies have described the role of RE in memory and executive functions, considerably fewer have examined its involvement in arousal and attentional states. While, as discussed earlier, the medial/midline thalamus was originally viewed as “non-specific” thalamus based on its generally widespread (and undifferentiated) projections to the cortex, a series of studies by Groenewegen and colleagues (Groenewegen et al., 1991; Berendse and Groenewegen, 1991; Groenewegen and Berendse, 1994; Van der Werf et al., 2002; Groenewegen and Witter, 2004) showed that each midline nucleus targeted very specific and separate forebrain structures, primarily those of the striatum and prefrontal cortex. This argued for distinct, perhaps unique functions for each of the midline nuclei. Following on this, Van Der Werf et al.

(2002) proposed that midline thalamic nuclei, as a whole, were critically involved in processes of arousal and attention – with individual groups of midline nuclei exerting distinct effects on behavior through the modulation of attentional processes. For instance, they proposed a role for the dorsal midline thalamus (mainly PV and PT) in

‘viscerosensory awareness’ based on its connections with the cortex and input from visceral-related structures, as well as one for RE/RH in ‘polymodal sensory awareness’ given its widespread subcortical and cortical connections spanning various modalities: sensori-motor, affect and memory.

In a series of studies, Schiff and co-workers (Schiff and Plum, 2000; Schiff and

Purpura 2002; Kobylarz and Schiff, 2005; Schiff et al., 2007; Schiff, 2008; Liu et al.,

2015) provided evidence to conclude that the medial or midline thalamus (MT) was

“uniquely positioned to control dynamic patterns of activation across cerebral networks” and thus, was an important source of “excitation” to the cortex in normal states and disorders of consciousness. As was pointed out, the reciprocal connections of MT with several cortical structures supports the influence of MT on large scale networks intimately associated with the planning and execution of behavior. They showed that the activity of populations of MT neurons was directly correlated with levels of vigilance in the rat (Liu et al., 2015), indicating a key role for MT cells in regulating and maintaining states of consciousness.

Gent et al. (2018) recently examined the properties of MT neurons during non-

REM sleep (NREM) and during sleep-wake behavioral transitions. They found that tonic optogenetic stimulation of MT produced arousal from NREM, while ‘burst-like’ optogenetically driven firing of MT cells resulted in slow wave (synchronous) activity in the cortex -- without awakening the animal. The activity of MT neurons significantly differed from that of sensory nuclei of the thalamus, as the latter group did not show the same sleep associated changes as MT cells (Gent et al., 2018).

Purpose and Dissertation Organization

The main purpose of this dissertation is to further investigate and define the role of the RE in spatial navigation, cognitive processes and arousal by examining its effects on spatial working memory, behavioral flexibility, and arousal/attention.

Previous work has demonstrated that RE is a key regulator of prefrontal- hippocampal interactions, however, its influence on working memory tasks at moderate

(seconds) to long (minutes) delays remains poorly described. I sought to investigate the role of RE in spatial working memory using a WM task that is dependent on the interactions of the HF and the mPFC. The first experiment (described in Chapter 1), examined the effects of the reversible inactivation of RE with muscimol (or procaine) on a delayed alternation working memory T-maze task using randomized moderate to long delays (30s, 60s or 120s). Using the same paradigm, I sought to examine the role of RE in executive functions by assessing the ability of rats to correct their behavior following failed trials on this task.

To further elucidate the contribution of RE in SWM and behavioral flexibility, a chemogenetic technique was used to either globally inactivate RE/RH or their terminal projections to main target sites -- using the same behavioral paradigm as in experiment 1.

Accordingly, in Chapter 2, an inhibitory DREADD (hM4Di) was expressed in RE/RH neurons (1) to inhibit their activity via the systemic delivery of clozapine-n-oxide (CNO); and (2) to selectively inhibit RE/RH terminals in the medial prefrontal cortex (mPFC) or the dorsal and ventral hippocampus via the local (intracranial) delivery of CNO at these sites. In Chapter 3, I sought to examine the role of RE/RH neurons in arousal/attentional processes by examining the activity of RE neurons across the sleep-wake cycles in order

to better understand their contribution in cognitive-executive functions such as SWM. To accomplish this, I recorded and then described the electrophysiological properties of RE single units across sleep (SWS/REM) and active wake states in freely behaving rats – with special attention to fluctuations in activity to changing levels of arousal (W, SWS,

REM).

The dissertation focused on the following three aims:

Aim 1: Investigation of the role of RE in spatial working memory and behavioral flexibility by transiently inactivating RE with muscimol and procaine while rats perform a T-maze delayed alternation working memory task at moderate to long delays (30-60-120s).

Aim 2: Examination of the effects of reversible inactivation of RE/RH with

DREADDs to inhibit RE/RH neurons, or selective RE terminal fibers distributing to the mPFC, dHF or vHF, to assess their role in SWM and behavioral flexibility using the same task as in experiment 1.

Aim 3: Examination of the activity of RE cells (single units) across sleep (SWS-

REM) and active wake states in freely behaving rats to determine how changes in activity across states of vigilance may contribute to RE related cognitive and executive functions.

CHAPTER 1: INACTIVATION OF NUCLEUS REUNIENS IMPAIRS SPATIAL

WORKING MEMORY AND BEHAVIORAL FLEXIBILITY IN THE RAT

Executive Summary

The hippocampal formation (HF) and medial prefrontal cortex (mPFC) play critical roles in spatial working memory (SWM). The nucleus reuniens (RE) of the ventral midline thalamus is an important anatomical link between the HF and mPFC and as such is crucially involved in SWM functions that recruit both structures. Little is known, however, regarding the role of RE in other behaviors mediated by this circuit. In the present study, we examined the role of RE in spatial working memory and executive functioning following reversible inactivation of RE with either muscimol or procaine.

Rats were implanted with an indwelling cannula targeting RE and trained in a delayed non-match to sample spatial alternation T‐maze task. For the task, sample and choice runs were separated by moderate or long delays (30, 60, and 120 s). Following asymptotic performance, rats were tested following infusions of drug or vehicle.

Muscimol infused into RE impaired SWM at all delays, whereby procaine only impaired performance at the longest delays. Furthermore, RE inactivation with muscimol produced a failure in win‐shift strategy as well as severe spatial perseveration, whereby rats persistently made re‐entries into incorrect arms during correction trials, despite the absence of reward. This demonstrated marked changes in behavioral flexibility and response strategy. These results strengthen the role of nucleus reuniens as a pivotal link between hippocampus and prefrontal cortex in cognitive and executive functions and

suggest that nucleus reuniens may be a potential target in the treatment of CNS disorders such as schizophrenia, attention deficit hyperactivity disorder, addiction, and obsessive‐ compulsive disorder, whose symptoms are defined by hippocampal‐prefrontal dysfunctions.

APPENDEX A: Viena, T.D, Linley, S.B. & Vertes, R.P. (2018) Inactivation of nucleus reuniens impairs spatial working memory and behavioral flexibility in the rat,

Hippocampus, doi:10.1002/hipo.22831, Reproduced with permission from Wiley, Inc.

CHAPTER 2: CHEMOGENETIC INACTIVATION OF NUCLEUS REUNIENS

OR RE FIBERS SELECTIVELY DISTRIBUTING TO THE MPFC AND

HIPPOCAMPUS: MEMORY AND GOAL DIRECTED ACTIONS

Introduction

As discussed, nucleus reuniens (RE) is strongly reciprocally connected with the medial prefrontal cortex (mPFC) and the hippocampus (HF) – structures essential for spatial working memory (SWM) and executive functions (see General Introduction). As described in Chapter 1, the reversible inactivation of RE with muscimol impairs choice accuracy and produces severe perseverative responding in a delayed non-matching to sample (DNMS) SWM task. These findings indicate that RE serves a critical role in

(spatial) working memory and behavioral flexibility.

Working memory, as first described by Baddley (1986) is the ability to temporarily hold task relevant information for subsequent choice behavior. Spatial working memory, in particular, is a process wherein relevant visuo-spatial information is gathered, encoded, held or ‘worked on’, and finally retrieved to guide goal directed behavior (Dudchenko, 2004; Myroshnychenko et al., 2017).

One way of assessing SWM in animal models is the delayed non-match to sample

(DNMS) paradigm (Munn, 1950; Dudchenko, 2004). DNMS tasks consist of a sample or encoding phase, a delay when information is held/manipulated, and a choice phase wherein a plan of action is executed (Dudchenko, 2001; Dudchenko, 2004;

Myroshnychenko et al., 2017). The T-maze is one of the most used paradigms to test

SWM and has been adapted for the DNMS task (for a review see Dudchenko, 2004).

The role of hippocampal and prefrontal interactions in the delayed nonmatch to sample paradigm

Several studies have demonstrated that interactions between the HF and the mPFC are important in SWM tasks (see General Introduction). Importantly, each structure serves unique roles which likely mediate different aspects of the task. For example, the HF is known for its direct role in various aspects of spatial behavior

(Buzsaki & Moser, 2013). Notably, neurons of the HF code for spatio-temporal properties thereby providing a neural representation of the spatial environment for spatial navigation (Okeefe & Conway, 1978; Eichenbaum et al., 1999; Kraus et al., 2013; Byrne,

2017). The dorsal hippocampus (dHF) has traditionally been viewed as the predominant

HF region for controlling spatial behavior. Specifically, lesions or inactivation of the dHF significantly alter SWM in various delayed matching (DMTS) and non-matching

(DNMS) tasks. While the ventral hippocampus (vHF) has been thought to be more fundamental to non-spatial than spatial processing, there is growing evidence for its involvement in SWM (Moser et al., 1993; 1995; Bannerman et al., 1995; Floresco et al.,

1996, 1997; Moser and Moser, 1998; Levin et al., 1999; 2002; Umegaki et al., 2001;

Arthur and Levin, 2002; Ferbinteanu et al.,2003; Broadbent et al., 2004; Pothuizen et al.,

2004; Wang and Cai, 2008; Loureiro et al., 2012). For instance, place cells have been identified in the vHF -- although considerably fewer in number and less spatially tuned than those of the dHF (Jung et al., 1994; Poucet et al., 1994; Kjelstrup et al., 2008; Royer et al., 2010; Komorowski, et al. 2013; Strange et al., 2014).

The mPFC also plays a critical role in working memory (WM). Early studies identified units of the PFC that discharged at enhanced rates across the delay period of

WM tasks. This increase in activity was presumed to be the mechanism that supported the temporary storage of information necessary for successfully executing a WM task

(Goldman-Rakic, 1996).

Several studies have implicated the mPFC in various functions that are essential in SWM (see Chapter 1 for review). For instance, during WM tasks, mPFC cells fire in response to path trajectories and trajectory changes, self-referential processing, context and contextual changes, strategy selection and implementation, as well as reward and reward expectancy (Jung et al., 1998; Jung et al., 2000; Rushworth et al., 2011; Yang et al., 2014; Rodgers and Deweese, 2014, Ito et al., 2015; Griffin et al., 2015; Euston et al.,

2012). As has been demonstrated, the disruption of the mPFC significantly alters WM, particularly in tasks requiring delayed responding (Shaw and Aggleton, 1993; Floresco et al., 1997; Seamans et al., 1995; Dias and Aggleton, 2000; Churchwell et al., 2010; Wang and Cai, 2006; Euston et al., 2012). For example, Dias and Aggleton (2000) demonstrated that mPFC lesions severely disrupted reversal learning on a delayed non-match to place

(DNMTP) task, leading the authors to conclude that the failure to reverse behavior based on changing circumstances involved an inability to switch strategies and/or shift attention. Therefore, one of the roles of the mPFC seems to be the ability to adjust to changing circumstances when current strategies are no longer useful, allowing animals to quickly adapt their behavioral responses (Brown and Tait, 2015).

The simultaneous engagement of mPFC-HF during SWM tasks suggests that the exchange and binding of spatial and contextual information between these two structures

likely supports the encoding, maintenance and retrieval of memories essential for performing the task (Cohen et al., 1996; Miller, 1999; Komorowski et al., 2013). Along these lines, several reports have shown that hippocampal-prefrontal power, coherence and phase locking are reliable predictors of correct choices in SWM tasks (Jones and Wilson,

2005; Hyman et al., 2010; Sigurdsson et al., 2010; O’Neil et al., 2013; Spellman et al.,

2015; Hallock et al., 2016; Myroshnychenko et al., 2017; Tamura et al., 2017; Liu et al.,

2018).

RE as a mediator of HF-mPFC interactions in cognition

As described, RE is reciprocally connected to the mPFC, dHF and vHF, and in the absence of direct mPFC-HF projections, RE is the major route from the mPFC to the HF

(Vertes et al., 2015b). As such, RE is strategically posited to modulate the flow of information between the mPFC and the HF during goal directed actions (Hoover and

Vertes, 2007).

The present experiment used a chemogenetic technique to compare the effects of inactivation of RE or selective RE projections to the mPFC, dHF and vHF using a modified T-maze DNMS task (Yoon et al., 2008; Zhang et al., 2012; Viena et al., 2018).

Based on our previous findings (Viena et al., 2018), it was hypothesized that the suppression of RE with inhibitory DREADDs would produce significant impairments in

SWM, flexible behavior and strategy selection. We further predicted that inhibition of

RE projections to the dHF, vHF or the mPFC would yield differential effects on behavior.

It was hypothesized that the suppression of RE projections to mPFC or dHF would produce a disruption of SWM and perseverative behavior, whereas inhibition of RE fibers to the vHF would also alter SWM, but not result in perseverative responding.

Finally, a further goal of the study was to assess changes in latencies to respond at critical junctures of the T-maze (start box, choice point) with increases in latency possibly indicative of failures to properly encode or retrieve information, while reductions of latency suggestive of impulsivity or lack inhibiting inappropriate responses.

Materials and Methods

Subjects

Subjects were adult male Long Evans rats (n=11) weighing between 275-350 grams upon arrival (Charles Rivers, Wilmington MA). Rats were housed in pairs in a climate-controlled colony on a 12 h light/dark cycle (lights on at 7 AM), during the experimental task. All animals had access to enriching materials in their homecage throughout the experiment. Upon arrival, rats were habituated for a minimum of 7 days, where food and water was provided at libitum. Rats were then placed on food restriction during pre-training and maintained at 85-90% their free feeding weight for the remainder of the experiment. Following surgery, rats were housed singly, and allowed to recover for 7 days, where they received food ad libitum, before being placed back on a food restricted diet, re-trained and tested on the task. The experiment was approved by the

Florida Atlantic University Institutional Animal Care and Use Committee and conform to all federal regulations and National Institutes of Health guidelines for the care and use of laboratory animals.

DREADDs Technique

DREADDs is a relatively new chemogenetic technique that allows for the reversible and less invasive manipulation of neuronal signaling (for review Campbell &

Marchant, 2018). In contrast to other techniques, DREADDs are highly specific, do not

destroy fibers of passage, and effects are reversible, usually returning to baseline activity within a few hours (Guettier et al., 2009; Alexander et al., 2009; Whissell et al., 2016).

The hM4Di DREADD receptor stem is genetically modified from the endogenous human

M4 muscarinic receptor, such that these endogenous receptors mutate and lose their ability to bind to their native ligand (acetylcholine), binding instead to the synthetic drug clozapine-N-oxide (CNO; Rogan and Roth, 2011; Wess et al., 2013). When CNO binds to the hM4Di receptor, a Gi signaling pathway results in the inhibition of cells by preventing the release of adenylyl cyclase and opening potassium channels which consequently, hyperpolarize neurons (Armbruster et al., 2007; Rogan and Roth, 2011).

Surgery and Viral Vector Delivery

Following completion of pretraining protocol on the DNMS task (see below), rats underwent surgery for the delivery of the chicken beta-actin promoter (CGA) serotype 9 adeno-associated virus (AAV) vector to induce muscarinic receptor 4 neuron-specific expression of the inhibitory DREADD hM4Di fused with mCherry (AAV9-CAG- mCherry-2a-hM4Dnrxn; Penn Vector Core/Addgene, Philadelphia, PA) into nucleus reuniens. Rats were anesthetized with isoflurane (4-5% induction, 1.5-3% maintenance) and locally anesthetized with a s.c. injection of 0.5% marcaine along the incision line.

Rats were then placed in a stereotaxic frame and an incision was made to expose the skull and holes drilled over the midline thalamus to deliver the vector in RE. Drilled holes over the prefrontal cortex and hippocampus were also made to position guide cannulas in these structures (22 gauge for single or 26 gauge double guides; Plastics One, Roanoke,

VA). Rats were injected with 0.3-0.4 ul of the DREADDs viral vector into RE/RH

(single stereotaxic coordinates: -2.4 mm AP; 0.0 mm L; 7.2 mm DV; bilateral: -1.8/-2.8

AP; 0.0 mm L; 6.7/7.3 mm DV). The virus was loaded into a Hamilton syringe and delivered via pressure injection at an average flow rate of 65 nl/minute. Following delivery, the syringe remained in place for 20 minutes to allow for diffusion and to avoid upward suction of the virus when removing the syringe. Next, rats were implanted with bilateral cannula guides in the prelimbic cortex of the mPFC (stereotaxic coordinates:

+2.7/2.8 mm AP; +/- 0.6 mm L; 2.5 mm cannula guide, DV) and bilateral cannulas over the dorsal CA1 (coordinates: 3.7/4.0 mm AP; +/- 2.2 mm L; 2.5 mm cannula guide, DV)

(n=5), or the ventral CA1 (coordinates: -5.8 mm AP; +/- 5.5 mm L; 5.0 mm cannula guides, DV) (n=5) of the HF (see Figure 6A-B). Stainless steel bone screws were anchored to the skull, and the headgear was affixed to the skull using dental cement

(Lang Dental, Wheeling, IL). Dummy cannulas were inserted in the implanted guide cannulas extending 1 mm beyond their tips. Rats were given one s.c. injection of

Buprenorphine SR (0.6 mg/kg) following surgery as an analgesic. Rats were given 7 days to recuperate before being returned to a food restricted diet and resuming training on the task. Vehicle or drug infusions did not begin until 21 days post-surgically to allow adequate time for viral expression of hM4Di throughout RE and its terminals.

Drugs and microinfusion procedures

For intraperitoneal injections (i.p.), CNO (Tocris, Minneapolis, MN) was dissolved at a concentration of 5mg/mL in normal sterile saline (sodium chloride; Vedco,

MO). Systemic injections were administered at a dose of 5mg/kg. Rats were given a minimum of three (3) saline injections prior to data collection to habituate them to the i.p. injection procedures.

Microinfusions of CNO or vehicle occurred once animals reached 70% accuracy

Figure 6 Set up for chemogenetically (with DREADDs) inactivating RE and its main targets. A) Silencing of RE with an inhibitory DREADD injected in RE (in purple) and activated by i.p. injections of CNO (5mg/kg) 30 minutes before the testing session. B) Placement location of bilateral cannulas in major cortical targets of RE (mPFC, dHF or vHF). Post-expression of the DREADD AAV viral vector in RE and RE projections, CNO (1mg/mL) was delivered (on separate testing days) in cannulated regions 30 minutes before testing.

Figure 6 A

on over two consecutive ‘mock’ infusions (for detailed procedures, see Chapter 1).

During infusions, rats were lightly restrained by the experimenter, internal cannula infuser(s) were inserted into the guide cannulas of rats and connected to a Hamilton syringe via a polyethylene tube. The syringe was loaded into a syringe pump (Model 22

Syringe Pump, Harvard Apparatus, Holliston, MA) and the drug or vehicle was delivered via microinfusion. Rats received vehicle or CNO (1 mg/mL) microinfusions bilaterally into the mPFC (0.5µl) and the dorsal or ventral hippocampus (0.75µl) at a flow rate of

0.25 µl/min. The order of both i.p. and infusions were randomized and counterbalanced across rats. Rate and dose for microinfusions of CNO were estimated based on previous work using this technique in dorsal hippocampus, orbital cortex and ventral tegmental area (Ge et al., 2017; Lichtenberg et al., 2017; Mahler et al, 2014).

Apparatus

The behavioral testing took place in a black wooden T-maze (described in detail in experiment 1). The maze was elevated from the floor using a table and located inside a room which had a black curtain covering the left side wall (remaining walls were white), a video camera on top, an entry door to the front, and a fire alarm box to right, all of which served as prominent and constant contextual cues. A single halogen lamp, located underneath the maze, provided illumination to the room. As in Experiment 1, during all training and testing sessions, white noise played continuously in the background.

Behavioral Testing

Prior to training, rats were acclimated to the testing room and maze for 3 days.

Next, animals were trained on a spatial alternation task, first with no delay, then on

‘fixed’ delays which increased in 30s increments, using a DNMS paradigm (as described

in detail in experiment 1). In brief, rats were given 10 daily trials, with each trial consisting of 2 runs (sample and choice), whereby correct performance required alternating goal arms between runs. Rats were allowed to make as many correction runs

(up to a maximum of 10 corrections) until the correct directional response was achieved.

Once performance reached a level of 70% correct for 3 consecutive training sessions, rats proceeded to the next training stage. Once rats completed all the ‘fixed’ delay training stages, they moved to randomized delayed alternation sessions.

Randomized delayed alternation sessions consisted of 10 trials with 30s or 120s delays, presented at random order, but equally represented in each testing session. Following a return into the startbox after a successful choice run, the arms were re-baited and the next trial began – yielding an averaged intertrial interval (ITI) of 10 seconds.

Upon reaching performance above 70% on 2 consecutive days on the randomized delayed alternation stage, rats underwent the surgical procedures described in section

Surgeries and Viral Delivery. Randomized delayed alternation re-training sessions resumed following the postsurgical recovery period. Upon reaching 70% or above on 2 consecutive days (but not before 21 days post-surgery) experimental injections or microinfusion procedures began. Rats were given vehicle or CNO in all target sites (i.p. or intracerebrally) in a within-subjects design. Following an experimental condition, rats were required to reach 70% or above asymptomatic performance for at least 2 consecutive days (non-infusion days), followed by mock testing, before undergoing another experimental manipulation. See Figure 7 for a diagram depicting the experimental design of this experiment.

Figure 7 Schematic representation of procedures for testing the effects of DREADD injections in RE using the delayed non-match to sample (DNMS) T-maze task. A) Shows the sequence of steps for training (top) and testing (bottom) rats on DNMS T-maze task. B) Shows sample and choices phases of the delayed alternation T-maze task. Rats made a free choice of the two arms (sample), and were then required to choose the opposite arm on the next run for reward (correct choice). Red ovals represent the areas in which latency was examined and compared across testing sessions. Bottom red oval, time to leave start box (SB); middle, deliberation time at T-junction (INT); and top (left or right), time elapsed from the start of the run to reward-cup arrival (GA).

Figure 7

Following the completion of all conditions, rats were deeply anesthetized with

250 mg/kg Euthasol and transcardially perfused with 4% paraformaldehyde, brains removed, fixed for another 24-48 hours and processed for histological verification of the

DREADD infection and cannula placement.

Data Analysis

Coding of behavioral data

Testing sessions were recorded on a HD camera (Amcrest Technologies), affixed to the ceiling of the room, and video files of experimental testing sessions were extracted for analysis. Two independent coders (inter rater reliability Cronbach's alpha = .97) analyzed rats’ testing sessions for performance measures and latency to respond during all trials (sample and the choice runs only).

Performance measures

Choice accuracy was used as a measure of SWM performance and was expressed as the percentage of correct responses (% correct trials/attempted trials) per delay during the testing session. Spatial perseverative responses (perseveration errors) were used as an index of behavioral flexibility, and defined as the inability to alternate to the opposite arm during correction runs following a choice run error (WM error). Videotaped latency responses during sample and choice runs at the startbox (SB), intersection (INT) and goal area (GA) were defined as follows: (1) initiation of action (SB latency) or the time taken to exit the SB at the beginning of each run once the door opened and the trial/run was initiated; (2) deliberation (INT latency) as the time required to make a directional response at the intersection of the T-maze; and (3) time to reward (GA latency) defined as the time to reach the goal/reward once the trial/run initiated (SB door opened). Latency

was measured in seconds. In addition, the speed of the rat (meters/second) was calculated using the measures previously described, to determine if treatment influenced locomotor behavior. Speed was computed by adding SB latency and INT latency, subtracting that from GA latency, and then comparing it to the distance between the SB door and the reward at the GA area (1.57 meters). All data was first entered into Microsoft excel prior to statistical analysis.

Immunohistology

Brains were transferred to 30% sucrose solution for 48 hours, then cut coronally using a sliding microtome into 50 µm coronal sections, collecting every six sections into series of tissue which were: (1) processed using a Nissl body stain to identify cannula placement in all brain areas of interest; (2) processed for anti-fluorescent red protein immunofluorescence to identify DREADD expression in both RE and RE terminal sites; and (3) immunofluorescence procedures to visualize the spread of DREADD viral infection (mCherry stained neurons) and neuronal bodies (NeuN). One set of free floating sections was mounted in chrome-alum gelatin slides for cresyl violet staining processing, microscopic visualization and photographic capture.

For anti-fluorescence red protein immunofluorescence, free-floating sections were first incubated for 30 minutes in 1% sodium borohydride mixed with 0.1 M Phosphate

Base (PB), then washed 3 times for 5 minutes. Next, tissue was blocked for one hour in a

0.5% BSA solution. After blocking, the slices were added to a 0.1% BSA solution that contained the primary antibody rabbit fluorescent protein (RFP 1:1000; Rockland

Antibodies). Tissue was covered in tin foil and incubated for 24 hours at room temperature and then another 24 hours at -4 degrees. On day 2, brain sections were

washed with PB (3 times for 5 minutes) and then placed in secondary biotinylated goat anti-rabbit antibody (1:500; Vector Labs). Secondary incubation lasted 2 hours and subsequently, tissue was washed with PB (3 times for 5 minutes). Sections were then incubated for 1 hour in the tertiary antibody biotinylated horse anti-goat antibody (1:500;

Vector Labs). Following the incubation period, tissue was washed with PB and then place in the ABC incubation solution (Vector Labs) for 1 hour. After this period, the tissue was washed and exposed to a chromagen solution until the developed stain was visible, but never longer than 3 minutes. After this, free floating sections were immediately washed in PB several times to stop the reaction. Upon completing the process, brain sections were ordered and mounted on gel-coated slides and coverslipped with VectaShield® Mounting Medium (Vector Laboratories, USA) for later visualization and photographic capture of sections using a digital microscope.

For mCherry and NeuN immunofluorescence, free-floating sections were processed and blocked as described in the previous paragraph. After blocking, the slices were added to a 0.1% BSA solution that contained the primary antibodies: 1) rabbit fluorescent protein (RFP 1:1000; Rockland Antibodies) and 2) mouse NeuN (1:500;

Millipore). Tissue was covered in foil and incubated for 24 hours at room temperature.

On the following day, tissue was washed 3 times for 5 minutes and transferred into a solution that contained the secondary antibodies for a period of 2-4 hours incubation, covered in foil and at room temperature. Secondary antibody solution contained Dylight

594 anti-Rabbit/Dylight 488 anti-mouse antibodies (1:3000) in 0.1% BSA solution.

Following this period, sections were washed 3 times for 5 minutes with PB, mounted on gel-coated slides and coverslipped with VectaShield® Mounting Medium (Vector

Laboratories, USA). Fluorescence was visualized with an epifluorescence and/or a high magnification confocal microscope.

The estimated viral proportion of DREADD infected neurons was obtained using the computer application ImageJ (NIH software) to do a cell count and averaging of dually labeled immunofluorescence cells (mCherry labeled/NeuN labeled x 100) from three representative (but similar) anterior, mid and posterior RE level sections (100 um by 100 um area) from all subjects. Verification of cannula placement in mPFC, dHF and vHF was conducted on brain sections stained with cresyl violet and visualized under a microscope.

Statistical Analysis

A 2 (delay) x 2 (experimental condition) mixed design analysis of variance was conducted on all measures unless otherwise noted. Homogeneity of variances, assessed by

Levene’s test of homogeneity, was not significant across conditions, unless otherwise indicated. Independently sampled t-tests were performed post hoc if significance was found. Estimates of effect size were calculated using partial eta squared (η2). Because of the relatively small sample sizes and violation of the homogeneity of variance in the RE > dHF and RE > vHF groups, a non-parametric test was used (Kruskal Wallis, KW).

One way analysis of variance were used to calculate differences among treatment groups related to win-shift failures and latencies at specific areas of the T-maze across brain areas of interest. Cases in which homogeneity of variance was significant, a Welch test was used to determine significance. As described above, non-parametric testing was performed in groups with small sample sizes (n ≤5). All calculations that were below the

0.05 alpha level (p value) were considered significant. Statistic tests were computed using

IBM SPSS 24.0 virtual application.

Results

DREADDs viral expression and histological verification

Ten rats were included in the study. Figure 8A-C depicts photomicrographs of

RE neurons, DREADD labeled cells (using mCherry) (A-C) or NeuN positive cells (GFP labeled) (middle panels of B-C) or double labeled cells (right panel of C) at low (B) and high levels of magnification (C). The bar graph of Figure 8 illustrates the mean percent of

DREADD cells relative to the total number of NeuN+ neurons in a subset of rats (n=2).

At rostral, mid and posterior levels of RE, the mean percent of DREADD labeled cells was 61% (SD=27.61), 64% (SD=25.09), 55% (SD=14.96), respectively. One rat was excluded from this analysis due to a lack of adequate numbers of DREADD labeled cells in RE. As shown in Figure 9, mCherry Anti-sera (red fluorescent protein) reacted brain tissue was used to identify RE DREADD infected fibers innervating dHF (dCA1), vHF

(vCA1) and mPFC (IL/PL, layers 1 & 5/6) in order to verify their viral expression following injections of the AAV viral vector into RE.

Figure 10 depicts cannula placements for all cases in target areas of interest

(mPFC, dHF, vHF). The cannula in the mPFC was placed in the ventral mPFC above or extending into prelimbic cortex (PL; n = 7; Figure. 10A). In a similar manner, the cannulas in the dorsal hippocampus were aimed for the dorsal portion of CA1, extending into CA1 pyramidal layer (n = 5; Figure10B). The cannula in the ventral hippocampus were placed at the mid-portion of the septotemporal axis of HF, within the ventral CA1 (n

= 5; Figure 10C).

Figure 8 Immunoreacted sections through the thalamus showing DREADD-infected cells in RE. A) Photomicrograph through the ventral midline thalamus showing the expression of the DREADD, AAV9-hM4Di-mCherry, in cells of the RE/RH. B-C) High magnification confocal micrographs, at two levels of magnification, showing single and double fluorescently labeled RE cells using mCherry and NeuN. D) Chart depicting the percentage of NeuN cells that were also labeled with mCherry at anterior, mid and posterior levels of RE.

Figure 8

Figure 8 continued

Figure 9 Distribution of mCherry-immunolabeled RE fibers in target structures. Photomicrographs mCherry labeled RE fibers in the A) medial prefrontal cortex and B) the dorsal and C) ventral hippocampus.

Figure 9

Figure 10 Nissl stained and associated schematic coronal sections showing cannula placements in the mPFC, dHF and vHF. A-C) Left: photomicrographs of cresyl stained transverse sections showing the locations of cannula placements in the mPFC, dorsal CA1 of HF, and the ventral CA1 of HF. Right: parallel schematic sections showing the locations of cannula placements for the subsequent CNO microinjections in these structures. Each symbol represents one microinjection.

Figure 10

Chemogenetic inhibition of RE alters choice accuracy across delays but does not produce spatial perseveration

As described (see Methods), rats were pretrained on the DNMP T-maze task prior to surgery to criteria performance and then re-trained post-surgery. The mean number of sessions to reach criteria for pre-training 23.00 (SD = 13.43) and the mean number of sessions post-surgically was 19. 00 (SD = 7.54). A mixed repeated measures ANOVA was conducted to compare the main effect of treatment and delay, and the interactions between treatment and delay on choice accuracy. Treatment included 2 conditions

(vehicle and CNO) and delay consisted of two conditions (30s, 120s).

An examination of the effect of treatment on choice accuracy on the T-maze task revealed a significant difference between vehicle (M = 80.00, SD = 4.69) and CNO (M =

54.27, SD = 5.32) injections, F(1, 14) = 13.15, p < 0.01, η2 = .48. Post hoc analysis using independent t-tests revealed a significant difference between vehicle (M = 88.89,

SD = 10.54) and CNO (M = 71.43, SD = 19.52) at 30 sec delays, t(1,14) =2.30, p < 0.05, and a significant difference between vehicle (M = 71.11, SD = 28.48) and CNO (M =

37.14, SD = 13.80) at 120 sec delays, t(1,14) = 3.136, p < 0.01 (see Figure 11). A main effect of delay in choice accuracy was found between the two delays, F(1, 14) = 14.19, p

< 0.01, partial η2 = .50, indicating that the longer delays (M= 54.13, SD = 5.88) were more difficult than the shorter delays (M = 80.00, SD = 3.80). There was no significant interaction between the treatment and delay on choice accuracy, F(1, 14) = 1.43, p >

0.05, partial η2 = .09.

With respect to perseveration on the task (or failure to correct responses following incorrect choices), there was not a statistically significant difference in perseverative

Figure 11 Effects of DREADD suppression of RE on choice behavior in DNMS T-maze task at two delay times. The inactivation of RE with systemic injections of CNO produced deficits in mean choice accuracy on the DNMS task at 30s and 120s delays (*p < 0.05; ** p < 0.01 respectively). Red dots, CNO microinjections; blue dots, vehicle microinjection.

Figure 11

errors between vehicle and CNO manipulations, F(1, 13) = .03, p > 0.05, η2 = .002.

Additionally, there was no main effect of delay on perseverative errors, F(1, 13) = 1.18, p > 0.05, partial η2 = .08, and no significant interaction between the treatment and delay on perseverative errors, F(1, 13) = .09, p > 0.05, partial η2 = .007 (see Figure 13A).

Chemogenetic inactivation of RE terminal projections to the vHF impairs working memory performance at intermediate delays

The silencing of the RE > vHF pathway showed a statistically significant difference of treatment on choice accuracy during 30s delays, H(1) = 4.615, p < 0.05,

(vehicle infusions mean rank = 7.50, CNO infusions mean rank = 3.50). On the other hand, there was no statistically significant difference of treatment on choice accuracy during 120s delays, H(1) = 1.224, p > 0.05, (vehicle infusions mean rank = 6.50, CNO infusions mean rank = 4.50). Figure 12A illustrates the mean choice accuracy of all individual cases following 30s and 120s delays between treatment conditions.

With respect to perseveration errors (see above), the silencing of RE > vHF pathway did not yield a statistically significant difference of treatment on perseverative errors during 30s delays, H(1) = 1.000, p > 0.05, or for the 120s delays, H(1) = 1.000, p >

0.05. Mean perseverations errors for all cases across delays and treatments conditions are shown in Figure 13B.

Chemogenetic inhibition of the RE > dHF pathway does not alter choice accuracy, or produce perseveration

Unlike the suppression of RE terminals in the vHF, chemogenetic silencing of RE projections to the dHF with CNO did not significantly altered choice accuracy across 30s delays, H(1) = .267, p > 0.05; or across 120s delays, H(1) = .886, p > 0.05, when

compared to vehicle infusions. Figure 12B shows the mean choice accuracy during 30s and 120s delays between treatments for all cases following the inhibition of the RE>dHF pathway. In like manner, perseveration errors were not significantly different between vehicle and CNO infusions into RE terminal ends in the dHF across 30s delays, H(1) =

.000, p > 0.05, or during 120s delays, H(1) = 1.000, p > 0.05. The mean number of perseveration errors during 30s and 120s delays across treatments for all cases is shown in Figure 13C.

Chemogenetic inhibition of the RE > mPFC pathway does not impair choice accuracy nor produce perseveration

Similar to the previous results, a repeated measures ANOVA revealed no main effect of treatment on mean choice accuracy following the chemogenetic inactivation of the RE > mPFC pathway, F(1, 12) = 1.13, p > 0.05, partial η2 = .09 (see Figure 12C), and no main effect of delay on mean choice accuracy between the two delays, F(1, 12) = .16, p > 0.05, partial η2 = .01. Additionally, there was no significant interaction between treatment and delay on choice accuracy, F(1, 12) = 1.46, p > 0.05, partial η2 = .11.

With respect to perseverative errors, there was no statistically significant difference in mean perseveration errors between the vehicle and CNO groups, F(1, 12) =

.375, p > 0.05, partial η2 = .030, (see Figure 13D); and no significant difference in mean perseverative responses across delays, F(1, 12) = .30, p > 0.05, partial η2 = .024. Finally, there was no significant interaction between the delay and treatment for perseveration errors, F(1, 12) = .30, p > 0.05, partial η2 = .024.

Figure 12 Effects of DREADD suppression of RE terminals on choice behavior in the DNMS T-maze task. A) The inactivation of RE terminals at the ventral hippocampus with CNO significantly impaired choice accuracy at 30s delays (top; p < 0.05) but not for during 120s delays. B-C) The inactivation of RE terminals at the dorsal hippocampus (B) or the mPFC (C) with CNO did not significantly alter choice accuracy for 30s or 120s delay times. Red dots, CNO microinjections; blue dots, vehicle microinjections.

Figure 12

Figure 13 Effects of DREADD suppression of RE as well as RE terminals to targets on perseverative behavior. A) The inactivation of RE with systemic injections of CNO did not produce perseverative responses at 30s or 120s delays. B-D). The inactivation of RE terminals at the ventral hippocampus (B), the dorsal hippocampus (C) or the mPFC (D) with CNO did not induce perseverative responses at 30s or 120s delays (all p’s > 0.05). Red dots, CNO microinjections; blue dots, vehicle microinjections.

Figure 13

A B

C D

Systemic CNO inactivation and CNO inhibition of RE terminals in the mPFC, dHF and vHF did not produce win-shift errors between trials

Win-shift errors were analyzed to investigate whether inhibitory DREADDs altered spatial alternation strategy (alternating following completion of a choice run and the start of a new sample run) following the systemic inactivation of RE/RH or the inactivation of RE terminals to target areas. Analysis of variance yielded no significant effect of treatment, Welch F(1, 10.20) = 3.99, p > 0.05, following the chemogenetic silencing of RE/RH neurons. There was also no significant effect of treatment on win- shift errors following the inactivation of the RE > mPFC pathway, F(1, 13) = .08, p >

0.05); the RE > dHF pathway, (H(1) = .600, p > 0.05; or the RE > vHF pathway (H(1) =

.046, p > 0.05).

Inactivation of RE terminals in the dHF or vHF, but not in the mPFC, alters latencies to initiate actions and time to reach reward site on the DNMS task

Latency at specific phases of the DNMS task were compared across delays and run types (sample/choice) following infusions of CNO or vehicle in RE terminal ends in the mPFC, dHF or vHF.

T-maze start box (SB)

Latencies to exit the SB at the start of the sample or choice runs following inactivation of RE terminals in the mPFC, dHF and VHF areas was examined.

Sample runs

As shown in Table 1, the mean latency to leave the start box for sample runs following CNO infusions in RE terminals in the dHF (M = 3.44; SD = 2.85) significantly decreased (more rapid exit) compared to vehicle controls (M = 10.02; SD = 8.94) for 30s

delays (p < 0.05) but not for 120s delays. By marked contrast, CNO infusions RE fibers in the vHF (M = 11.02; SD = 8.92) resulted in a significant increase in latencies (slower exit) compared to controls for 120s delays (M = 3.72; SD = 3.19; p < 0.05), but not for the 30s delays. No significant differences were found for intracranial injections of CNO in RE terminal ends in the mPFC for 30s or 120s delays (all p > 0.05).

Choice runs

No significant differences in latencies to leave the SB during the 30s delay choice runs was found between infusions of CNO and vehicle in RE fibers in the dHF, mPFC or vHF groups (all p > 0.05); however, infusions into the vHF approached significance (p

=.07).

Following 120s delays, the time to leave the SB after infusions of CNO in RE terminals to vHF (M = 8.10; SD = 16.47) were significantly longer than for vehicle controls (M= 2.16; SD = 1.70; p < 0.05). On the other hand, the time to leave the SB was not significantly different across treatment groups for either the mPFC or dHF (all p >

0.05).

T-maze junction

The time spent making a directional decision at the junction of the maze was not altered by CNO infusions in RE terminals for either sample or choice runs or across any of the delays in any of the target areas (all p > 0.05) when compared to controls. Notably however, rats spent considerably longer time at the T-maze junction following CNO injections in vHF at 120s delays for sample and choice runs, but this difference was not statistically significant (Table 1).

Table 1 Mean latencies at critical points in the T-maze during performance of the DNMS task following CNO microinjections in RE target sites. Mean latencies (and SDs) in seconds to exit the startbox, deliberation at the T-junction and to reach the reward area of the T-maze for sample and choice runs at 30s (top) and 120s (bottom) delays with inactivation of RE fibers to the mPFC, and to the dorsal or ventral HF (left to right), compared to controls. Values in bold were statistically significant. Values in blue highlight were not statistically different, but approached significance (p. <.051-.07).

following task

DNMS

the of during performance maze -

pointsT the in

Mean latencies at critical at latencies Mean sites. in target RE CNOmicroinjections 1 Table

T-maze goal area

Once a run was initiated, the time to arrive to the reward (GA) was assessed across the sample and choice runs for both 30s and 120s delays.

Sample runs

At 30 second delays, the latency to arrive at the reward after CNO infusions in RE terminal ends in the dHF was significantly shorter than for controls (p = 0.05). On the other hand, no significant differences between saline and CNO infusions were seen in RE terminal fibers in mPFC (p > 0.05) or vHF (p > 0.05) for 30s delays -- although for the terminals in vHF there was a trend (p = .07) toward significance.

At 120 second delays, the latency to arrive to the reward following CNO infusions of RE terminal projections in the vHF (M = 15.74; SD = 15.92) was significantly longer than for controls (M = 6.74; SD = 3.51; p < 0.05). No significant differences in the time to arrive at the goal area were seen at 120s delay for the mPFC or the dHF (all p > 0.05)

Choice runs

The time to arrive at the reward during 30s delay choice runs did not significantly differ across any target area (all p > 0.05). However, the time to reach the GA was twice as long for infusions in RE fibers in the vHF compared to controls at 30s, but this was not statistically significant (p = 0.07).

Choice runs at 120s delays following infusions of CNO in RE terminal ends in the mPFC or dHF did not statistically differ (p > 0.05). Similarly, at 120s delays, inactivation of RE terminals in the vHF did not significantly differ in latency, however this effect was a marginally significant (p = 0.054).

Locomotive speed

No significant differences in running speed (motor deficits) were found between

CNO and saline infusions during sample or choice runs across 30s or 120s delays for any of the RE projections to the mPFC, dHF or vHF (all p > 0.05; Table 2). Accordingly,

CNO infusions did not alter the speed in which rats navigated the maze when compared to vehicle controls.

Table 2 Mean speeds during performance on the DNMS T-maze task following CNO injections in RE target sites. Means (and SDs) locomotor speeds (meters/s) for sample and choice runs at 30s (top) and 120s (bottom) with CNO inactivation of RE fibers to the mPFC, and to the dorsal or ventral HF (left to right), compared to controls. As shown, there were no significant differences in running speeds across conditions or RE target sites (p > 0.05).

maze task following CNO injections following task injections maze CNO -

Table 2 Table T during DNMS the on performance speeds Mean sites. target RE in

Discussion

Chemogenetic inactivation of the nucleus reuniens (RE) significantly impaired choice accuracy in a delayed non-match to sample (DNMS) T-maze task. Following delays of 30 or 120 seconds, i.p. CNO, but not vehicle, injections significantly impaired performance in treated rats. Additionally, the selective inactivation of RE terminals to the ventral hippocampus (vHF), but not those to the dorsal hippocampus (dHF) or medial prefrontal cortex (mPFC), disrupted delayed spatial alternation at intermediate (30s) delays.

Successful execution of the DNMS task requires subjects to retain the directional information of the previous run (sample) during the delay and then, use that information to choose the opposite arm in the subsequent (choice) run for reward. As such, the task relies upon spatial working memory (SWM). Chemogenetic suppression of RE activity, as well as inhibition of the RE > vHF pathway, altered performance on this this task -- indicating a deficit in SWM. The results are consistent with those of Griffin and colleagues (Layfield et al., 2015; Hallock et al., 2016) which showed that muscimol inactivation of RE impaired working memory performance at 30s delays in a similar

DNMS task, as well as the findings of Experiment 1, which demonstrated that the reversible inactivation of RE with muscimol impaired SWM performance at 30, 60, and

120s delays. Similarly, Hembrook and Mair (2011) reported that RE lesions produced marked deficits on a delayed 8-arm radial maze (RAM) task following delays of 1, 5, 10 and 30 minutes.

RE > ventral hippocampal pathway: critical for SWM and goal directed behaviors

Given the mounting evidence supporting a role for RE in SWM via actions on the mPFC and HF, we examined the effects of selective inactivation of RE projections to vHF on the DNMS task and found that inhibiting RE terminals to the vHF significantly altered choice accuracy at moderate (30s), but not long delays (120s) delays. In line with these findings, various studies have demonstrated a role for the vHF in SWM (see

General Introduction). In particular, the vHF is critical in encoding task relevant locations

(goals/contextual cues), synchronizing the vHF and mPFC in the gamma range, and mediating the synchrony (in the theta band) between the dHF and mPFC (O’Neil et al.,

2013; Spellman et al., 2015). For instance, O’Neil and colleagues (2013) using a SWM paradigm similar to the present experiment, found that the pharmacological inactivation of vHF with muscimol significantly altered theta synchrony between the dHF and mPFC and choice performance. Spellman et al. (2015), using an optogenetic approach to suppress vHF projections to the mPFC, further demonstrated that the vHF is necessary in the synchronization of vHF-mPFC gamma oscillations and accurate performance on a

DNMS task.

While the vHF projects to, and hence can directly affect, the mPFC (see Figure 2), there are no direct return projections from the mPFC to the HF (Vertes et al., 2015b).

Therefore, feedback from the mPFC to HF is likely routed indirectly through the RE.

Specifically, the vHF provides the mPFC with a representation of goal associated cues

(reward, context), the mPFC in turn, evaluates the information, utilizes it during the delay period and then, in cooperation with RE, provides the vHF with updated task details that appear essential for guiding subsequent actions.

In addition to routing spatial contextual information from the mPFC to the vHF,

RE also affects stimulus-driven motor behaviors. Chemogenetic suppression of RE terminals to the vHF increased the latency to exit the start box (SB) and also increased the time to reach the reward. However, this was only significant in the sample

(encoding) and choice (retrieval) phases of the task at 120s delays. The observed increases in latencies were not the result of impaired locomotion, as comparisons of running speeds with CNO and vehicle across conditions and phases of the task showed no differences (p > 0.05). Instead it appeared that rats were not prepared to initiate the trial once it began. Additionally, this group of rats paused for several seconds before making a directional response at the T-junction and made more non-goal directed movements

(rearing) along the arms of the maze. The increased latency at both sample and choice runs at the start box, suggests that these changes in behavior are related to attention, which potentially can affect the encoding and retrieving phases of the task. Along these lines, RE/RH receive important input from arousal-related systems (see General introduction) and recent studies have highlighted the role of RE in attentional processes such as visuospatial attention and attentional set shifting (Prasad et al., 2013; Cholvin et al., 2013; Linley et al., 2016; Cholvin et al., 2017).

Another possibility could be that CNO infusions into the vHF produced an anxiogenic response. The vHF is well known for its role in emotional regulation (see

General Introduction). Along these lines, Padilla-Coreano et al. (2016) recently demonstrated that optogenetic inactivation of vHF terminals to the mPFC altered anxiety behaviors and spatial representation to aversive outcomes. Likewise, lesions of the mPFC in rats, specifically IL/PL, alters the latency to escape aversive situations (Jinks

and McGregor, 1997). RE has also been postulated to be a cognitive-emotional regulator

(Davoodi et al., 2011; Xu and Sudhof, 2013; Kafetzopoulos et al., 2017). Taken together these studies indicate that information sent from the mPFC to the vHF via RE is likely associated with the regulation of cognitive-affective behaviors during SWM tasks. The impairments could, of course, be also the result of a combination of both: inattention and increased anxiety.

Contrary to our expectations, CNO-induced inhibition of RE terminals at the mPFC or dHF did not disrupt SWM performance. Whereas the reasons for this lack of effect at these sites is presently unclear, it is known that RE inputs to the dHF are relatively sparse compared to those to the vHF (Vertes et al., 2015b). It is therefore possible that inactivation of this pathway requires a much greater suppression of RE inputs to dHF to see impairments. Interestingly, while we did not observe deficits in choice accuracy on the task, CNO infused at the dHF significantly reduced the time it took rats to exit the SB and the time to reach the reward. These results are similar to those of Prasad et al. (2013) which tested the role of RE in attention using a 5-CSRTT task and found that RE lesioned rats were notably faster in collecting a reward (reduced latency) and made premature responses (lack of impulse inhibition). Our findings also indicate an inability to suppress a prepotent well established motor responses

(impulsivity), but this did not affect task performance. Therefore, it can be concluded that possibly one function served by the mPFC-RE-dHF pathway is the inhibition of motor responses.

Chemogenetic inactivation of RE or RE projections to the hippocampus or prefrontal cortex did not produce spatial perseveration in the DNMS task

In contrast to the first experiment (Viena et al., 2018), systemic injections of CNO did not produce perseverative errors (p > 0.05). Specifically, immediately after rats made an error during the choice run (SWM error), they were given the opportunity to repeat the run (without delay) to correct their mistake. Unlike the perseveration seen with muscimol, the DREADD treated rats quickly corrected their behavior after a choice error, with most of them alternating to the opposite arm on the first try (first correction run).

Since the experimental conditions for the muscimol and DREADD studies were the same, it seems very likely that muscimol may exert a greater suppressive effect on RE than do DREADDs. In line with this idea, Walsh et al. (2017) demonstrated that RE neurons in vitro exhibit diverse firing patterns and responded differently to biophysical and pharmacological manipulations. As previously discussed, hM4Di receptors hyperpolarize cells via the Gi signaling cascade. Muscimol, on the other hand, is a

GABAA receptor agonist that inhibits neuronal activity by the opening of chloride channels, producing an influx of chloride and driving the cell to more hyperpolarized levels. Although the final mechanisms of action may be similar for both agents, it appears that muscimol is much longer acting than DREADDs (Martin and Ghez, 1999, Mahler et al., 2014), and muscimol is also effective in suppressing calcium transients, while

DREADDs are less effective (Cichon & Gan, 2015). It is also possible that muscimol, compared to DREADD-CNO, affected a much larger population of RE neurons (Carter et al., 2013; Miao et al., 2015; Chang et al., 2015). On this particular, we demonstrated that the viral expression of DREADDs in RE only affected ~60% of all cells (see Results),

thus although CNO suppression may have been robust, it did not encompass the totality of cells in RE.

One limitation of this experiment is the relatively small sample size of the RE to dHF and vHF groups. Although we observed clear differences between CNO and vehicle conditions, the relatively small number of rats is probably the reason for not reaching statistical significance in some of the measures of interest. A further limitation of this study was the lack of proper DREADD controls. The concern has been raised that with systemic CNO injections high concentrations of CNO can accumulate over time converting to clozapine, which potentially can activate endogenous brain receptors

(Gomez et al., 2017). We partially addressed these concerns by using low to moderate doses of CNO (5mg/kg systemic; 1mg/mL infusions), and further our experiments were of short duration (45-60 minutes), well below the suggested time range for the safe interpretation of results using CNO (Mahler and Aston-Jones, 2018). Additionally, our intracranial infusions bypass the metabolic process of CNO, as it is delivered into a restricted region of the brain. Finally, several studies to date have demonstrated the lack of unforeseen behavioral effects following CNO microinjections directly into the brain

(Mahler et al, 2014; McGlinchey and Aston-Jones, 2018; Ge et al, 2017; Mahler and

Aston-Jones, 2018).

Conclusions

The present experiment demonstrated that global chemogenetic inactivation of RE neurons with CNO or the inactivation of RE terminals to vHF, significantly altered SWM performance in a DNMS task that engages the mPFC-HF circuit, further supporting the role for RE as an important link in communication between the mPFC and HF in decision

making and goal directed actions. Additionally, we found that CNO-suppression of RE fibers to the vHF increased the latency to initiate motor actions and reach the reward, while CNO silencing of RE axons to dHF produced the opposite effect – significantly decreasing the time it takes the rat to initiate actions on the task. These findings point to a role for RE in the regulation of attentional, affective (emotional reactivity), and inhibitory motor control.

Psychiatric disorders, such as schizophrenia, have been related to alterations in the mPFC-HF network in which RE plays a pivotal role. The present findings add to current knowledge and further implicate RE in cognitive domains beyond that of working memory.

RE involvement in cognitive, attentional, motor and affective processes makes it an ideal therapeutic target for the treatment of various mental health related-illnesses.

CHAPTER 3: CHARACTERIZATION OF UNITS OF NUCLEUS REUNIENS

ACROSS SLEEP AND WAKING STATES

Introduction

Role of the thalamus in the modulation of the cortical EEG during wake-sleep states

The thalamus regulates cortical EEG activity during wake and sleep states. In the waking state, cells of the specific and “non-specific” thalamus discharge tonically (tonic mode) producing high frequency low amplitude activity across the cortical mantle: beta and gamma activity (Fanselow et al., 2001). Conversely, in sleep, thalamocortical activity enters a ‘burst’ mode as the result of a weakened brainstem arousal input and the inhibitory actions of the reticular thalamic nucleus (RT) on thalamocortical neurons

(Fanselow et al., 2001; Ward, 2013), giving rise to delta/theta waves in the cerebral cortex.

Activity of thalamic nuclei in sleep and arousal

The activity of thalamic nuclei across the wake-sleep states was first explored in cats. Classical studies indicated that the overall activity of thalamic cells was similar during waking (W) and REM sleep, but significantly reduced their discharge in slow wave sleep (SWS). For instance, cells of various thalamic relay nuclei fired at high tonic and regular rates in W/REM, while in SWS they fired in a slow bursting pattern separated by long silent periods of activity (Hubel, 1960; Lamarre et al., 1971; Dormont, 1972;

Benoit and Chataigneir, 1973; Steriade et al., 1987; Sherman and Guillery, 2001).

In sleep, several mechanisms regulate thalamocortical oscillations: (1) the interactions between RT, the cortex and other thalamic nuclei; (2) the intrinsic properties of RT and thalamocortical (TC) neurons; and (3) ascending “arousal” influences from the brainstem on the thalamus. At the neuronal level, fast Na+/K+ currents mediate the tonic firing (depolarization) of thalamic cells during wakefulness (Steriade et al., 1993;

Sherman and Guillery, 2001). Conversely, during sleep the inhibitory influence of the RT coupled with the voltage gated membrane conductances (IT, Ih currents) of TC cells result in the generation of short bursts of action potentials (Steriade et al., 1993; Sherman and Guillery, 2001; Coulon et al., 2012). Specifically, during NREM sleep RT-induced hyperpolarization of TC cells de-inactivates low threshold T-type Ca2+ channels and upon rebound, activation generates a bursting discharge in these neurons (Jahnsen and

Llinas, 1984; Steriade et al., 1993; Sherman and Guillery, 2001; Ramcharan et al., 2000;

Brown et al., 2012). In this regard, Lewis et al. (2015) recently demonstrated that optogenetic stimulation of RT dramatically reduced firing rate of TC neurons, switching them from a tonic to a bursting mode that produced slow wave oscillations in the cortex.

Ascending excitatory input from the brainstem to the thalamus is another important factor influencing the discharge modes of TC neurons. Cholinergic, noradrenergic (NE), and serotonergic (5-HT) release to the thalamus induces the depolarization of thalamocortical neurons via the inactivation T-type Ca2+ channels, which switch TC cells from a bursting mode in NREM to a tonic mode of discharge in waking. The removal of excitatory brainstem influences on TC neurons, coupled with RT actions on them (see above), again, generates bursting activity of TC neurons that

resumes slow wave activity in the cortex during NREM sleep (Steriade et al, 1993a;

Steriade et al, 1993b; Steriade et al 1993c; Sherman and Guillery, 1996).

Whereas early reports focused on the role of “first order” thalamocortical cells in the regulation of the cortical EEG in sleep/wake states (Steriade et al., 1987;Contreras and Steriade, 1995; Sherman and Guillery, 1996; Sterman, 1996; Sherman, 2001), more recent reports have shown that “higher order” thalamic neurons, including those of the medial/midline thalamus, exhibit similar discharge profiles across sleep and waking states as first order TC neurons – and thus also significantly contribute to state-dependent changes (Sheroziya and Timofeev, 2014; Liu et al., 2015). For instance, Liu et al. (2015) demonstrated that medial thalamic stimulation not only gave rise to a transition from sleep to waking (with associated changes in the cortical EEG), but converted a drowsy state to one of attentive wakefulness.

Supporting this Gent et al. (2018) recently showed that cells of the midline central medial (CM) nucleus in behaving mice fired in advance of the cortical EEG/behavioral transition to waking and optogenetic stimulation of CM, but not sensory thalamic nuclei, aroused sleeping mice. The foregoing findings indicate a critical involvement of the midline thalamus in attentional states, possibly mediating the effects of attention on cognition – in line with purported role for the midline thalamus in arousal and attention

(Van der Werf et al., 2002; McKenna and Vertes, 2004; Vertes, 2006; Vertes et al., 2007;

Hoover and Vertes, 2011, Saalmann, 2014).

Role of nucleus reuniens in sleep-waking and attention

As discussed, several recent reports have described the involvement of RE in

SWM and executive functions (Dolleman-van der Weel et al., 2009; Cholvin et al., 2013;

2017; Prasad et al., 2013; Prasad et al., 2017; Linley et al., 2016). Currently, it is unclear the degree to which alterations in attention contribute to cognitive deficits produced by the inactivation of RE. This, in part, was examined here. In addition, no study has examined the firing properties of RE cells across waking and sleep states.

Accordingly, experiment 3 examined the discharge properties of RE neurons across the sleep (SWS-REM) and wake cycles in freely moving rats. Following acute sleep deprivation procedure (6-18 hours), single RE unit activity was assessed as rats cycled through waking, NREM and REM states. Sleep-wake states were identified with reference to the hippocampal and cortical EEG and behavioral observation. It was hypothesized that RE cells would shows distinct changes across sleep-wake states, possibly like other thalamic cells (McCarley et al., 1983; Steriade and Llinas, 1988;

Steriade et al., 1989; Sherman and Guillery, 1996; Sherman, 2001) discharging at low rates and in bursts in NREM and at high tonic rates in waking and REM sleep. Finally, the heightened activity of RE cells in wakefulness could suggest a role for them in attentional processes – to thus mediate the effects of attention on cognitive-executive functions linked to nucleus reuniens.

Materials and Methods

Subjects

Thirteen male Sprague-Dawley rats (Charles River, Wilmington, MA;

Harlan/Envigo, Dublin, VA) weighing 275-325 grams at the beginning of the experiment were housed individually in a humidity controlled environment (vivarium) with a 12 hour light/12 hour dark cycle. Access to food and water was available ad libitum. Enrichment materials were available in the cages of rats throughout the experimental period and were

rotated at regular intervals. Following a week acclimatization to the vivarium, rats were surgically implanted with electrodes in midline thalamus and hippocampus for the recording of single units and the hippocampal EEG. All procedures were in accordance with federal regulations, the National Institute of Health (NIH) guidelines for the care and use of laboratory animals and the Florida Atlantic University Institutional Animal Care and Use Committee standards.

Surgical Procedures

Rats were anesthetized with isoflurane (4-5% induction, 1.5-3% maintenance), given Buprenorphine SR 0.6mg/kg for long term pain management, and the skin above the skull was locally anesthetized with a s.c. injection of 0.5% marcaine. Immediately after, rats were placed in a stereotaxic frame and an incision was made to expose the skull. Two burr holes were drilled over the midline thalamus in an anterior/posterior plane with ~1.0 mm separation (stereotaxic coordinates: -1.6/-1.7 and -2.6/-2.7 mm AP;

+0.4 mm L; 6.0 – 7.0 mm DV) for implantation of a single movable microelectrode containing two tetrodes targeting mid to rostral levels of RE. Next, a chronically fixed bipolar electrode aiming to the left CA1 of the dorsal hippocampus (coordinates: -3.5 mm

AP; -3.0 mm L; 3.0 mm DV) was implanted. Four clips and 2 bone screws were anchored to the skull to secure the entire microelectrode assembly. An EEG recording stainless steel (SS) screw was placed on the skull over the frontal cortex, while a grounding screw was attached to the skull over the parietal cortex. Clips, screws and electrodes were securely anchored using dental cement (Lang Dental, Wheeling, IL).

Loose skin tissue around the acrylic was glued using vetbond surgical glue (3M, St. Paul,

MN) and any large skin gaps were sutured and sealed with vetbond. Rats were given 7

days to recuperate before recording sessions began.

Electrode construction

A custom-built microelectrode was constructed for recordings of neurons in RE

(modified from Stackman et al., 2002), theta in dorsal hippocampus and the cortical EEG.

Each customized microelectrode consisted of two 10-pin omnetics connector arrays

(Omnetics, Minneapolis, MN) and a customized 3D printed microelectrode base that contained all of the microelectrode components. Figure 14 shows the components and assembly process.

For recordings units in RE, two tetrodes were constructed using four 25 μm diameter wires per tetrode (California Fine Wire, Grover Beach, CA) cut to a length (20 mm) and twisted at 25-30 revolutions in a clockwise direction and glued with vetbond

(3M, St. Paul, MN). Then, two 10-pin Omnetics arrays were glued together using loctite.

Tetrode wire end tips were then soldered to pins 1 through 8 of array 1, resulting in 8 recording channels. An additional wire was soldered to pin 9 of the array and served as an internal reference. For cortical EEG recordings a SS teflon coated wire (125um,

Medwire, Mt Vernon, NY) was soldered to a SS screw at one end and to the first pin

(recording channel 1) of array 2 on the other end. A bipolar twisted-pair of teflon-coated

SS wires (125 um, MedWire, Mt. Vernon, NY) separated by ~ 1 mm at their tips was used to record the hippocampal EEG from the dorsal hippocampus. The bipolar electrode wire tips were soldered to pins 2 and 3 (recording channels 2 and 3) of array 2. Pins 4-9 in array 2 were soldered together to shunt these channels from recording signals. Finally, a teflon coated SS wire was soldered to pin 10 on both arrays and attached to a SS screw

Figure 14 Components of customized drivable microelectrode assembly. Components of the microelectrode assembly: 3D printed base (top row) and construction stages of the drivable microelectrode (bottom row) for recording of RE units..

Figure 14

(ground screw) with the purpose of grounding the entire assembly.

Tetrodes were then placed into a 26-gauge SS tube to prevent the wires from bending when descending through the brain. Next, all the components were fixed to a 3D printed base (Tinkercad.com) using dental acrylic cement. Two drive screws (14mm long) attached to the 3D printed base were used to advance the microelectrode assembly to targets following implantation. The weight of the final electrode assembly was about

3.5 g. Finally, the conductance of the microelectrode wires was tested (prior to implantation) by passing DC current through the partially submerged tetrodes tips in a physiological saline solution (0.9% NaCl) at room temperature.

Apparatus

All unit screening and recording sessions were done in a Faraday cage ( 26”W x

30”L x 12” H) which was covered by a blue colored antistatic acrylic/pvc sheet. A black curtain surrounded the faraday cage, and an overhead DC lamp provided illumination. A color webcam (c920, Logitech, Mexico) was positioned inside the faraday cage, 79 cm above the base surface. A cue card was attached to left side inside wall of the cage. The cue card and the camera served as visual landmarks.

Procedures

Seven days post-surgery, rats were habituated to their transport, and acclimated to the testing room, recording apparatus and noise generator for 3 days prior to the start of recording sessions. Following this, rats were brought to the recording cage daily for unit screening. Rats were lightly restrained and connected to a microelectrode headstage (1x) attached to a tethered cable leading to a signal amplifier (Plexon, Dallas, TX). Rats were then placed in the enclosure, allowed to explore the arena and forage for chocolate treats

(20 mg; Bioserve, Flemington, NJ). During exploration, each channel was screened for the detection of units. When spikes were detected, the rat was sleep deprived (SD) on the following day. If no units were observed on any of the channels, the tetrodes were lowered in small increments deeper into the brain (~⅛ of turn = 75 um) until suitable cells were detected. This was done on the same or following day.

Sleep Deprivation (SD)

Rats were sleep deprived for a period of 6-18 hours using the inverted flower pot technique (Mendelson et al., 1974; Andersen et al.; 2004; Singh et al., 2008). A clay flower pot (6” diameter) placed ‘upside down’ was positioned in the middle of a circular container (painter’s bucket, dimensions: 12” diameter and 15” height). The container was filled with approximately two inches of water. Food and water were always available to the rats during SD period via a metal grid placed on top of the bucket (Figure

15A). Rats were monitored at regular intervals while in the container to ensure their safety. Sleep deprivation occurred during the evening hours and recordings commenced in the morning. Once rats were removed from the container, they were placed into their home cage for a period of 30 minutes prior to the start of recordings. During this period, food and water was available at libitum and rats were kept awake by tapping on their home cage. Each rat was recorded for about 2-3 sleep-wake cycles (2-4 hours).

Electrophysiological recordings

A 16-channel headstage (Plexon, Dallas, TX) connected to a tethered cable, carried signals to an amplifier (see above). Neuronal activity was pre-amplified (1000x), filtered (see below) and then digitized using a CED 1401 (Cambridge Electronic Design,

Cambridge, UK) acquisition and analysis unit. The behavior/movements of rats were

videotaped using a webcam, synchronized to the LFP recordings. Single spikes and EEG local field potentials (LFP) activity were recorded and saved to a personal computer for subsequent off-line analysis.

Single Units

Recorded signals were filtered (.03-10 kHz) using a Grass EEG amplifier (Grass,

Quincy, MA), then acquired and sorted using Spike 2 software (CED, Cambridge, UK) within each channel using an automated template program. Spike waveforms were sampled at a frequency of 22 kHz and recorded when signals exceeded the defined threshold.

Cortical and hippocampal EEG

The cortical EEG signal was filtered (.1-300 Hz) using a Grass EEG amplifier

(Grass, Quincy, MA). LFP was sampled (1 kHz) and stored using Spike 2 software.

Hippocampal EEG signals were filtered between 3 Hz (high pass) and 100 Hz (low pass) using a NPI (NPI, Farmingdale, NY) customized filter/amplifier, then sampled (200 Hz) and stored using Spike 2 software.

Histology

Upon completion of the experiment, rats were deeply anesthetized with isofluorane and a small electrolytic lesion was made by passing a small DC current (10-

20 μA, 3-10s; Peyrache et al., 2009) via the microelectrode to mark the location of the electrode tips, either at the end of the tract or at the levels in which units were identified.

Immediately after they were deeply anesthetized with Euthasol (250 mg/kg) and perfused with 400 ml of 4% paraformaldehyde. Brains were extracted and preserved in 4% paraformaldehyde for 24 hours and then in a 30% sucrose solution for another 48 hours

or until the brain sank to the bottom of the vial. Brains were then coronally sectioned at

50 um using a freezing microtome (Leica SM2000R) and mounted onto chrome–alum gelatin-coated slides. Brain sections were processed with cresyl violet, visualized under a microscope and photomicrographs were captured using a Nikon Eclipse E600 microscope and ACT software.

Data Analysis

Sleep Stage Coding Criteria

Vigilance stages were visually inspected and manually scored by a trained coder using the cortical EEG, hippocampal EEG, and observable behavior via pre-recorded videos -- following initial presorting and identification of vigilance states using a custom- made script in Spike2 (Ratsleepauto; Costa-Miserachs et al., 2003). The script uses the filtered hippocampal LFP signal (off-line) to automatically tag theta and non-theta periods of the hippocampal EEG, as well as changes in cortical EEG activity at 5s or 10s epochs to aid in the identification waking, NREM and REM states. State identification was done in Spike2, together with the video recording, based on the following polysomnographic criteria:

● Active wake (W): Desynchronized (or activated) low amplitude cortical EEG and

the visual confirmation of the rat in an alert state (engaged in exploration or

consummatory behavior) via the video recordings, supplemented by a hippocampal

EEG spectrum showing a predominant theta band (6-10 Hz; see Figure 15).

● Slow-Wave Sleep (SWS): Mainly high-voltage slow waves (1.5–4 Hz; delta) of the

cortical EEG with predominant non-synchronous (irregular) hippocampal EEG

activity and the rat immobile in a sleeping position with eyes closed.

● Rapid-eye-movement (REM) sleep: Low-voltage cortical EEG activity with

continuous hippocampal theta (6-10 Hz) and with a sleep posture and eyes closed.

Figure 15B shows an example of the observed behaviors across sleep-wake states, with corresponding hippocampal EEG, cortical EEG and hippocampal power spectrum.

Single unit waveform sorting

The first analysis was aimed at identifying the waveform of single units during the recording sessions. Spikes were identified based on amplitude and shape using a

Spike2 pre-defined waveform templates and automatically calculated threshold for each of the selected channels. Waveforms that were very similar to each other were combined into a single template. Clustering of waveform variance within templates was verified through principal component analysis (PCA) to ensure that detected spikes were distinct from other waveforms in single channels. Data was saved for later off-line analysis.

Spectral analysis of the cortical and hippocampal EEG

For each vigilance state, spectral analyses were made of the cortical and hippocampal EEG using Spike2 fast Fourier transform. Each period considered for analysis consisted of at least 10 s of a vigilance state, except in some instances of shorter

(5s) REM periods.

Separation of unit spikes across vigilance states

Stored spike waveforms were separated into groups based on their firing patterns, waveform, and discharge profile across S-W stages and autocorrelations. The firing rate of neurons across each sleep-wake cycle was analyzed with Spike2. Data was averaged, sorted and saved in Microsoft excel for further statistical analysis.

Figure 15 Method for sleep deprivation and behavioral/EEG signs of wake-sleep states. A) Diagram showing the flower pot method for sleep deprivation in rats. B) Left: video clips of the rat in waking and sleep (SWS and REM). Middle: simultaneous patterns of hippocampal (top) and cortical EEG activity in each state; and right: corresponding spectral analysis of the hippocampal EEG for each state. Note the prominent theta rhythm in waking and REM sleep.

Figure 15

Flower Pot Method A

1 sec

Firing discharge patterns

RE single units were assigned to low (L), low to moderate (LM) or high (H) groups based on rates of discharge by computing the 25, 50 and 75 percentile ranks of mean discharge rates using SPSS 24.0 (virtual app).

Spike waveforms

RE single cells were also grouped based on their waveforms type, as generated by

CED Spike2 template matching analysis program.

Separation by discharge profile across states

RE units were classified and then grouped in accordance to their discharge profile activity patterns across the W, SWS and REM states. Separation criteria was based on each unit’s highest to lowest mean firing rate activity across each of the vigilant states.

Statistical analysis

Analysis of mean firing rates was first performed using raw firing rates. As data was not normally distributed (as expected), analysis was repeated using z-normalized firing rates. Raw and normalized data produced the same pattern of results. Using the

Kruskal-Wallis (KW) nonparametric test, independent samples were taken and examined for differences in mean firing rates for each isolated cell across behavioral states (W,

SWS, REM). If significance was met, KW post hoc pairwise comparisons, using

Bonferroni corrections to p values, was used to identify statistical differences across groups. All statistical analysis were performed using SPSS version 25.0.

Results

Histological verification

Six rats (n = 6) were included in the analysis. Cases in which the anatomical location could not be verified (loss of implant; n = 3), equipment malfunction (video, headgear or high noise to signal, n = 2), or the tetrode track or tip was outside of RE (n =

2) were not included for analysis. The anatomical location of tetrode tracks and tips was determined by making small electrolytic lesions at electrode tips and noting electrode positions at the time of the recording (see Methods). Figure 16A-B depict representative

Nissl-stained sections showing electrolytic lesions in RE at different levels of the recorded units in one subject, and a lesion at the end of the recording tract in another.

Figure 16C-D depicts schematic sections through the ventral midline thalamus localizing the sites of the lesions for the six rats of the study. Figure 16E depicts a representative case of the electrode tract and its termination in CA1 of the dorsal hippocampus for recording of the hippocampal EEG. A total of 203 cells were recorded and analyzed across the behavioral states of active waking (W), slow wave sleep (SWS) and REM sleep, and were confined to mid to mid-rostral regions of RE.

RE discharge patterns and waveforms

The mean firing rate for all RE neurons during active wake (W) was 5.73 Hz (SD

= 4.99), for SWS 2.20 Hz (SD = 2.73), and REM sleep 4.67 (SD = 4.37) (Figure 17).

Table 3 summarizes the mean spontaneous firing rates of RE cells across the three behavioral states and within each state classifies cells according to low (L), low to moderate (LM) or high (H) rates of firing. On the whole, discharge rates of RE cells fluctuated across vigilant states, exhibiting maximum firing during wake periods,

Figure 16 Sites of unit recording in RE and slow wave recordings in the dorsal hippocampus. A-B) Photomicrographs of Nissl-stained transverse sections depicting the sites of recording (*) in RE for two rats (A and B); (C-D). Schematic transverse sections through the ventral midline thalamus showing the position of recording sites in RE for all cases that were analyzed (color). E) Nissl stained section showing the electrode tract and termination in CA1 of the dorsal hippocampus for recording the hippocampal EEG in a representative case.

Figure 16

Table 3 Mean discharge rates of RE cells units across sleep-wake states Discharge rates of recorded RE cells (203) across active waking, SWS and REM sleep: means, standard deviations, standard errors and minimum/maximum values. Neurons classified according to relative rates of firing in waking: low (L), moderate (M) and high (H) for each state.

wake states wake -

Table 3 Table of rates units sleep across discharge cells RE Mean

decreasing in NREM sleep and increasing in rate again in REM, but not to the level of

waking. Mean differences in firing rates of RE neurons across W, SWS and REM states was statistically significant (H(2) = 87.46, p < 0.001; mean rank: wake = 369.00, SWS =

213.33, REM = 332.97). Post hoc analysis using the Bonferroni correction revealed that mean firing rates were significantly different between W (M = 5.73, SD = 4.99) and SWS sleep (M = 2.21, SD = 2.73; p < 0.001), as well as between REM (M = 4.67, SD = 4.37) and SWS sleep (p < 0.001). On the other hand, no significant differences were found between W and REM (p > 0.05; see Figure 17). The same relationships were observed for each of the “firing type” groups (low, LM, and high; see Table 3).

In addition to classifying RE cells according to relative rates of discharge (Table

3), they were also categorized according to four distinct types of waveforms: negative positive (NP), negative (N), positive (P) and positive negative (PN) (Figure 18).

Approximately 32% of the cells were NP-type cells. These cells exhibited a prominent initial negative component followed by a large positive phase before returning to baseline. N cells represented 18% of the population and were characterized by a large negative phase with no positive component. P cells accounted for 29% of cells and displayed an initial large positive phase followed by a generally long duration but small amplitude negative component. Finally, PN units represented 22% of cells and consisted of a symmetrical positive-negative waveform which was similar to P cells but exhibited a larger amplitude negative component than did P cells.

Table 4 shows the means, standard deviations, and ranges of firing rate for cells classified by waveforms across W, SWS, and REM. There was no significant differences in mean firing rates of RE neurons grouped by waveforms for W, SWS, and

100

Figure 17 Mean firing rates of RE cells across waking, SWS and REM sleep. A comparison of the firing rates of RE neurons during active wake (W), slow-wave sleep (SWS) and REM sleep shows wake-sleep dependent changes. As a population, RE cells discharged at significantly higher rates in waking and REM sleep than in SWS. There were no significant differences in firing between active waking and REM sleep. Asterisks indicate significance level (p < 0.001)

101

Figure 17

102

Figure 18 Types of spike waveforms of RE units. Four distinct types of spike waveforms were found for RE cells. Negative- positive (NP) spikes were characterized by an initial negative component followed by a positive deflection (top left). Negative (N) spikes consisted of a negative component with no positive phase (top right). Positive (P) cells showed an initial large positive phase followed by small amplitude, long duration negative component. Positive-negative (PN) spikes exhibited a characteristic symmetrical positive-negative shape (bottom right).

103

Figure 18

104

Table 4. Mean discharge rates of RE cells classified by waveforms across sleep-wake states Discharge rates of RE cells classified by waveforms across active waking, SWS and REM sleep: means, standard deviations, standard of error and minimum/maximum values. No significant differences in mean firing rates were found for the different waveforms and across states (p > 0.05)

105

wake states wake -

4 Mean discharge rates of RE cells classified by waveforms across sleep across of waveforms rates classified by discharge cells RE Mean Table

106

REM sleep (W, H(3) = .174, p > 0.05; SWS, H(3) = 3.860, p > 0.05; REM, H(3) = 1.259, p > 0.05).

Comparison of discharge profiles of RE neurons across wake-sleep states—with wake- active (W) and REM-active (R) cells subdivided into W1, W2, R1 and R2 type cells

Next, RE cells were separated according to their discharge profile type across wake-sleep states. Five distinct cell types were identified: wake-active 1 (W1), wake- active 2 (W2), REM-active 1 (R1), REM-active 2 (R2) and slow wave sleep (S1) cells.

Table 5 compares the mean firing rates of RE’s W1, W2, R1, R2 and S1 cells across wake-sleep. W1 cells (123/203, 61%) had their highest firing rate during wake, reduced activity during REM sleep and lowest discharge rate in SWS (W>REM>SWS). W2 cells

(16/203; 8%) fired at highest rates during waking, but interestingly discharged at higher rates in SWS than in REM (W>SWS>REM). R1 and R2 cells fired at higher rates in

REM sleep than in the other two states. Whereas, as indicated, R1 and R2 cells were most active in REM sleep, R1 cells (51/203; 25%) discharged at higher rates in waking than in SWS (REM>W>SWS), while R2 cells (9/203; 4%) fired more in SWS than in waking (REM>SWS>W). Finally, S1 cells (4/203; 2%) fired maximally in SWS and at lowest rates in waking. As shown, however, S1 cells constituted a very small percentage of recorded neurons. Figure 19 A-C and Figure 20 A-C depict representative examples of RE single units firing discharge fluctuations across the W, SWS and REM states. Note the drastic changes in firing activity that occur between the S-W states. Figure 19D shows examples of W2 and R1 cell spike events recorded simultaneously across the vigilant states (W, SWS, REM) in one rat. Figure 20D demonstrates the activity of W1,

W2, R2 units recorded simultaneously across stages in another subject.

107

Table 5 Mean discharge for RE cells classified according to discharge profile across sleep-wake states. Discharge rates of RE cells classified according to sleep-wake profiles across active wake, SWS, and REM sleep: means, standard deviations, standard of error and minimum/maximum values. Five cell types were identified: wake-active (W1 and W2), REM-active (R1 and R2) and slow wave sleep (S1) cells (see text). Mean firing rates of W1, W2, R1, R2, S1 cells did not significantly differ in waking (p > 0.05), but mean rates significantly differed between cell types in SWS (p < 0.001) and REM sleep (p < 0.01).

108

wake states. wake

Table Table sleep profile across discharge to cells RE for classified discharge according Mean

109

The differences in mean firing rates of the five classes of RE neurons (W1, W2,

R1, R2, S1) during waking was not significant H(4) = 8.74, p > 0.05. On the other hand, differences in mean discharge rates of these groups were significant during SWS, H(4) =

22.52, p < 0.001; with a mean rank of 100.58 for RE W1 cells, 131.50 for W2 units,

82.64 for R1 units, 118.63 for S1 units and 171.28 for R2 type cells. KW post hoc pairwise comparisons (Bonferroni correction) showed that during SWS, the firing rate of

R2 units (M = 6.93, SD = 4.42) was significantly different from R1 units (M =1.53, SD =

2.03; p < 0.001) and W1 units (M = 1.91, SD = 2.09; p < 0.01). Additionally, the discharge rate of W2 units (M= 3.96, SD = 4.58) was significantly different from that of

R1 cells (M = 1.53, SD =2.03; p < 0.05).

Significant differences in mean firing rates for the five classes of cells was also observed in REM sleep, H(4) =15.19, p < 0.01, with a mean rank of 99.41 for W1 cells,

75.66 for W2 units, 111.18 for R1 units, 50.25 for S1 units and 155.22 for RE R2 cells.

Post hoc analysis showed that R2 type cells (M = 9.84, SD = 6.90) were significantly different than W2 cells (M = 3.36, SD = 4.45; p < 0.05) and S1 type cells (M = 1.30, SD

= 1.24; p < 0.05).

Patterns of RE activity during active W, NREM and REM sleep

The patterns of activity of RE cells differed across states. For instance, most cells fired tonically in active W and REM sleep, but with varying degrees of regularity.

Approximately 28% of cells exhibited a rhythmic bursting pattern of activity during

SWS, while ~ 71% showed no bursting pattern of discharge in SWS. As previously described, with few exceptions, RE cells fired at lower rates in SWS than in either W or

REM sleep: burst or non-bursting cells. (see Figures 19 and 20).

110

Figure 19 Discharge patterns of two types of RE cells across W, SWS and REM sleep. Representative samples of the discharge patterns of two types of RE cells across active waking, slow wave sleep and REM sleep (A-C). Top to bottom for A-C: counts of individual (single) spikes, hippocampal EEG, cortical EEG, RE LFP signal, time scale. D) enlarged spike counts; W (dark blue bars); SWS (green bars) REM (light blue bars). One type of cell, R1, fired at highest rates in REM, followed by W and SWS: REM>W>SWS. Another type of cell, W2, fired at highest rates in waking followed by SWS and REM: W>SWS>REM. Samples 10s. See also Table 5.

111

Figure 19

112

Figure 20 Discharge patterns of three types of RE cells across W, SWS and REM sleep. Shows representative samples of the discharge patterns of three types of RE cells across active waking, slow wave sleep and REM sleep (A-D). Top to bottom for A-C: counts of individual (single) spikes, hippocampal EEG, cortical EEG, RE LFP signal, time scale. D) enlarged spike counts; W (dark blue bars); SWS (green bars) REM (light blue bars). One type of cell, W1, (2 units) fired at highest rates in waking followed by REM and SWS: W>REM>SWS. A second type of cell, W2, fired at highest rates in waking followed by SWS and REM: W>SWS>REM. A third type of cell, R2, fired at highest rate in REM followed by SWS and waking: REM>SWS>W. Samples 10s. See also Table 5.

113

Figure 20

114

Discussion

The present study is the first to examine the electrophysiological properties of neurons of the nucleus reuniens across the sleep-waking cycles in freely behaving rats.

The results demonstrate a heterogeneity of rates and patterns of RE neural activity in sleep-wake states. We recorded 203 neurons in RE and showed that: (1) RE cells, on the whole, fire at a rate of about 6 Hz during waking, reduce in rate to ~ 2 Hz during SWS and increase again in rate to ~ 5 Hz during REM sleep; (2) based on their rate of activity during waking, RE neurons were divided into three groups, low (~1Hz), low to moderate

(~5Hz), and high frequency firing cells (~12Hz); (3) based on their differential rates of activity across active W, SWS, and REM sleep, cells were subdivided into five types ordered by the states in which they fired the most to the least; and (4) RE cells fired tonically in active W and REM sleep and a subpopulation discharged in bursts in SWS.

Although few reports have examined the discharge properties of RE neurons in behaving animals (and none across sleep-wake states) the rates and patterns (tonic discharge) of RE cells shown here in waking are very comparable to those described in several other reports in that state (Ito et al., 2015, 2018; Lara-Vasquez et al., 2016;

Jankowski et al., 2015; Gent et al. 2018). For instance, Gent et al. (2018) recently showed that cells of the midline thalamus in mice discharged spontaneously at 5.3 Hz during waking.

Diversity of RE activity across vigilant states

The majority of RE cells were classified as wake-active: W1 and W2 neurons.

Specifically, W1 cells, which comprised about 61% of the population, discharged at highest rates in waking, slowed in SWS, and increased in rate again in REM sleep to

115

about 80% of levels of waking. The discharge profile of these neurons is very similar to the properties of “non-specific” and relay neurons of the thalamus which typically fire at high (or maximal) rates in both waking and REM sleep and at low rates in SWS

(McCarly et al., 1983; Steriade and Llinas, 1988; Steriade et al., 1989; Sherman and

Guillery, 1996; Sherman, 2001; see Chapter 3 Introduction).

W2 cells (~ 8%), discharged in a similar manner to W1 cells in waking, but interestingly discharged more in SWS than in REM sleep thus showing a progressive decrease in firing rate across states: W > SWS > REM sleep. Notably, the discharge profile of these cells is similar to that of monoaminergic neurons of the brainstem, such as the dorsal raphe (DR) nucleus and the locus coeruleus (LC), that show this progressive decrease in discharge rate from waking to REM sleep (Hobson et al., 1975; McGinty and

Harper, 1976; Trulson and Jacobs, 1979; Aston-Jones and Bloom, 1981; Rasmussen et al., 1986; Shima et al., 1986; Levine and Jacobs, 1992; Rajkowski et al., 1994; Darracq et al., 1996; Nitz and Siegel, 1997; Gervanosi et al., 1998; 2000; Aston-Jones and Cohen,

2005; Carter et al., 2010). The DR and the LC, together with other brainstem nuclei, are critical components of an ascending arousal system for maintaining wakefulness

(Moruzzi and Magoun, 1949; Van der Werf et al., 2002; Schiff et al., 2007; Schiff, 2008,

2016; Monti, 2010a, 2010b; Berridge and Schmeiche, 2012; Monti and Jantos, 2014; Liu et al., 2015) and are also a rich source of input to RE and to other midline thalamic nuclei

(Jones and Moore, 1977; Sawchenko & Swanson, 1983; Peschanski and Besson, 1984;

Cornwall and Phillipson, 1988; Cornwall et al., 1990; Vertes, 1991; Waterhouse et al.,

1993; Krout et al., 2002; Van der Werf et al., 2002; McKenna and Vertes, 2004; Vertes and Linley, 2008, Vertes et al., 2015b; see General and Chapter 3 Introduction).

116

Accordingly, this brainstem input to RE may selectively drive a subpopulation of

RE neurons (W2 cells) during waking – which may then participate in RE-related processes such as arousal and attention (Foote et al., 1980; Snatucci et al, 1996; Aston-

Jones et al., 1999; Brown et al., 2002; Bouret and Sara, 2005; Carter et al., 2010; Rossetti and Carboni, 2005).

Whereas W1 and W2 cells discharged at highest rates in waking, the remaining

RE cells (31%) fired at their highest rates in either REM sleep (vast majority) or SWS sleep: R1, R2, or S1 cells. While the mean firing rate of R1, R2 or S1 cells did not significantly differ from that of W1/W2 cells during waking, differences were seen in

SWS or REM sleep. Specifically, two types of RE neurons were identified that fired at highest rates during REM sleep, R1 and R2 REM-active cells. R1 neurons (25%), like

W1/W2 cells, discharged at significantly higher rates in waking and REM sleep than in

SWS, but unlike W1/W2 cells fired at higher rates in REM than in waking which was the reverse of W1/W2 neurons. By comparison, R2 cells (4%) discharged in reverse order to

W2 cells (W > SWS > REM); that is, fired at highest rates in REM and at progressively lower rates in SWS and W. While few in number, similar types of REM-active neurons have been identified in other thalamic structures. For instance, Marks et al. (1981) described a population of LGN cells that fired in high frequency bursts in SWS and even more vigorously in REM sleep. These REM-active cells could be involved in the PGO spikes of REM sleep (Kasamatsu and Adey, 1973; McCarley et al., 1978; Marks et al.,

1981, 1999; Davenne, 1987; Steriade et al., 1989) or alternatively participate in the visual imagery of dreams/REM sleep (Steriade et al., 1989; Sprenger et al, 2010). The presently identified R2 (REM-active) neurons could serve functions unique to REM sleep as their

117

activity is strikingly similar to that of REM-ON cells. Given that stimulations of LDT and PPT neurons result in an increase of activity in midline thalamic cells (Pare and

Steriade, 1990; Steriade, 1996), and RE receives projections from brainstem nuclei known to harbor REM-ON cells (LDT/PPT; McKenna and Vertes, 2004), it is possible for some RE neurons to not only exhibit similar electrophysiological properties, but also serve similar functions.

One of the most prominent features of thalamic neurons during sleep is their characteristic rhythmical bursting activity in SWS which drives the slow oscillations of the cortex during non-REM sleep (Deschenes et al, 1984; Steriade et al., 1987;

McCormick and Pape, 1990; Weyand et al., 2001; Llinas and Steriade, 2006; Halassa et al., 2011; see Chapter 3 Introduction). While we identified some cells showing this bursting pattern of discharge in SWS (~28%), the majority of cells did not – firing at reduced but only sporadically in bursts during SWS. Although this pattern of activity slightly deviates from that which has been traditionally described in TC, this is not uncommon. For instance, recent investigations have reported the activity of sensory TC cells as mainly inhibited with irregular burst activity, while the activity of non-specific

TC is described as mostly active during SWS (Sheroziya and Timofeev, 2014; Honjoh et al., 2018). Although these studies were done in anesthesia-induced slow wave recordings in mice, they provide evidence of the differences in firing of TC cells during NREM sleep.

Conclusions

We showed that RE is composed of a diverse set of neurons that differ in activity across sleep-wake states. No previous report has examined the discharge properties of

118

RE neurons across sleep-waking states of any species. While some studies have described the activity of midline thalamic/RE cells in waking under various conditions (Ito et al.,

2015, 2018; Jankowski et al., 2014, 2015; Lara-Vasquez et al., 2016; Gent et al. 2018), virtually none have examined RE activity in sleep and especially REM sleep – due in part to the very short duration of REM episodes in rodents. We largely circumvented this issue by sleep depriving rats prior to recording using the inverted flower pot method

(Mendelson et al., 1974; Andersen et al., 2004; Singh et al, 2008) which then generally resulted in long bouts of REM sleep – from which to record.

Despite their heterogeneity, the majority of RE neurons fired at selectively high rates in active waking (W1 cells) suggesting a direct role for RE in states of arousal and attention (Van Der Werf et al., 2002; Vertes et al., 2007; Vertes, 2006; Pereira de

Vasconcelos and Cassel, 2014; Salalmann, 2014). In this regard, alterations of RE have been shown to disrupt behavior in various tasks requiring attention to salient environmental cues (Prasad et al., 2013, 2017; Varela et al., 2014; Linley et al., 2016, see

General Introduction). The deficits in SWM that we showed with the inactivation of RE

(studies 1 and 2) could, to a significant extent, involve the disruption of W1 neurons (or other RE cells), and the consequent loss of attentional influences on cognitive functions.

In summary, we described several types of RE cells with activity related to different stages of the sleep-wake cycle. This suggests that RE may serve various roles in waking and sleep. RE is reciprocally connected with the medial prefrontal cortex and the hippocampus (McKenna and Vertes, 2004, Hoover and Vertes, 2012; Vertes et al., 2006; see General Introduction) and is thus positioned to modulate the transfer of information between these structures. The content of information and its direction, and time course of

119

flow between mPFC and hippocampus via RE is likely to differ across sleep-wake states

(Buzsaki, 1986; Marrosu et al., 1995; Hasselmo, 1999; McNaughton et al., 2003; Sirota et al., 2003; Isomura et al., 2006; Molle and Born, 2009), indicating that RE may serve different, but perhaps complementary functions in sleep and waking.

120

GENERAL DISCUSSION

The purpose of the present work was to examine the role of nucleus reuniens (RE) of the midline thalamus in higher order functions. RE is a critical node in a hippocampal- prefrontal circuit, the epicenter of affective and mnemonic processes. In the first study, using a delayed non-match to sample (DNMS) T-maze task, the reversible inactivation of

RE via muscimol significantly impaired spatial working memory and produced pronounced spatial perseveration, a marker of behavioral inflexibility. These effects were presumed to result from the disconnection of hippocampal-prefrontal circuits mediated through RE. Next, to parse out the functional role of specific RE projections to key target structures, a chemogenetic approach was utilized. Rats were injected with an inhibitory

DREADD into RE and tested on the same DNMS task following the delivery of CNO, either systemically or intracerebrally, at RE or its targets sites in the dorsal (dHF) or ventral hippocampus (vHF), or the medial prefrontal cortex (mPFC). Reversible inactivation of RE or RE projections to the vHF produced impairments in spatial working memory (SWM), while inactivation of RE projections to the dHF or the mPFC did not produce significant deficits in SWM. Finally, single unit activity was recorded from RE in freely moving rats across sleep-wake states. The majority of cells exhibited a characteristic pattern of activity, discharging at high rates of activity in active waking and

REM sleep, and significantly lower rates in slow wave sleep (SWS) – with some showing a bursting pattern in SWS. The combined findings support a critical role for RE in working memory and executive functions.

121

RE as a conduit in the exchange of cognitive information between the hippocampus and the medial prefrontal cortex in spatial working memory

Previous work of the laboratory examined the anatomy of RE (Vertes, 2002;

McKenna and Vertes, 2004; Vertes et al., 2006, 2007; Hoover and Vertes, 2012). These studies not only described in detail the vast and rich input RE receives from the brainstem, diencephalon and cortex, but highlighted the unique and bidirectional connectivity linking RE with the HF and the mPFC (McKenna and Vertes, 2004; Vertes,

2006; Vertes et al., 2007; Hoover and Vertes, 2012).

RE contains a population of cells that project via axon collaterals to mPFC and

HF (Hoover and Vertes, 2012; Varela et al., 2014). While the hippocampus projects directly to the mPFC, there are no return mPFC-HF projections, and as such RE is ideally positioned to direct the flow of information from the mPFC to the HF, and hence mediate the coordinated actions of these two structures (Vertes, 2002; Vertes et al., 2006, 2007,

2015; Varela et al., 2014; Griffin, 2015).

Several reports have demonstrated that the HF and the mPFC act individually or, more commonly, in combination to SWM (Floresco et al., 1997; Lee and Kesner, 2003;

Jones and Wilson, 2005; Yoon et al., 2008; Churchwell et al., 2010; Churchwell and

Kesner, 2011; Colgin, 2011; Gordon, 2011; O’Neill et al., 2013; Preston and

Eichenbaum, 2013; Griffin, 2015; Sapiurka et al., 2016). Specifically, the HF and mPFC appear to function independently in working memory at brief (seconds) delays, but mutually interact at longer delays. Regarding the latter, Wang and Cai (2006) found that the bilateral inactivation of the HF, or the mPFC, or disconnection of these structures by unilaterally inactivating them in opposite hemispheres, impaired SWM on a delayed

122

alternation T-maze task. Similarly, Kesner and colleagues (Lee and Kesner, 2003;

Churchwell and Kesner, 2011) demonstrated that the HF and mPFC act in parallel to process spatial memory at 10s delays, but interact at 5 min delays when tested in a

DNMS radial arm maze task. These reports concluded that the maintenance and retrieval of relevant information during working memory tasks is dependent on the reciprocal interactions of the HF and the mPFC. It has further been shown that successful performance on SWM tasks is directly associated with synchronous oscillations between the HF and the mPFC, mainly in the theta band (Jones and Wilson, 2005; Siapas et al.,

2005; Benchenane et al., 2010; Hyman et al., 2010; Gordon, 2011; O’Neill et al., 2013;

Remondes and Wilson, 2013; Shin and Jadhav, 2016).

In the absence of direct projections from the dorsal HF to the mPFC or from the mPFC to the HF (see above), RE is well positioned to coordinate the activity of the HF and mPFC in SWM as well as other functions. Accordingly, several early reports have shown that RE lesions (or inactivation) disrupt HF-mPFC interactions and behavior

(Hembrook and Mair, 2011; Hembrook et al., 2012; Loureiro et al., 2012; Cholvin et al.,

2013; Hallock et al., 2013; Layfield et al., 2015).

For instance, Mair and collegues (Hembrook et al., 2012) demonstrated that the reversible inactivation of RE/RH with muscimol impaired performance on an operant chamber DNMP task, but not on a variable choice RAM DNMP task. Notably, while both of these tasks test for spatial working memory, the first one depends on the interactions of the mPFC and HF, while the variable choice RAM task is solely sensitive to hippocampal lesions and does not require the mPFC (Porter et al., 2000). In a similar manner, Hallock et al. (2013) examined the involvement of RE in a working memory

123

dependent and a non-working memory dependent discrimination T-maze task. In line with the findings of Hembrooke et al. (2012), they showed that RE was only involved in

SWM tasks requiring the combined actions of the HF and mPFC, but not those dependent on other brain regions such as the striatum. Finally, Layfield et al. (2015) investigated the role of RE/RH in delay dependent SWM task using a behavioral paradigm similar to the one used in the present experiments. They found that silencing RE did not affect performance in the continuous alternation version of the T-maze task, however, when delays were imposed (5s, 30s), SWM performance was significantly disrupted.

Electrophysiological studies in rodents have provided additional evidence for the involvement of RE in mediating HF-mPFC interactions and behavior. For example,

Hallock et al. (2016) using a DNMS T-maze working memory task recently demonstrated that inactivation of RE with muscimol not only resulted in poor SWM performance, but also disrupted synchronous oscillations between the dorsal HF and the mPFC during the task.

There is also an emerging role for RE in executive functions, such as attention, goal directed behavior, and behavioral flexibility. For instance, Dolleman–van der Weel et al. (2009) reported that RE lesions did not affect spatial memory in a water maze task, but rather led to a maladaptive search strategy, akin to a mPFC dysfunction in the task.

Prasad et al. (2013) found that RE lesions produced impairments in behavioral inhibition, as demonstrated by increases in premature (impulsive) responses on a 5-choice serial reaction time task (5-CSRTT). Finally, Linley et al. (2016) found RE lesions produced impairments in attention and reversal learning (a measure of behavioral flexibility) on an attentional set shifting task. These findings would indicate that RE not only serves a role

124

in SWM, which involves HF-mPFC interactions, but also regulates executive functions, most likely via RE actions on the orbital cortex (Young and Shapiro, 2011; Linley et al.,

2016; Wikenheiser and Schoenbaum, 2016).

In Chapter 1, we examined the effects of the inactivation of RE with muscimol or procaine on DNMS T-maze task, using three randomized delay times: 30, 60, or 120s.

We showed that the reversible inactivation of RE with muscimol impaired performance

(choice accuracy) at all delays, while procaine infusions only impaired choice accuracy at the longest delay. The inactivation of RE also produced a severe spatial perseverative behavior, in that rats made repeated entries into the incorrect and unrewarded arm of the maze -- when allowed to correct their behavior after incorrect choices at no delays. Thus, our results indicate a critical role for RE in SWM and behavioral flexibility.

As previously described, the RE had been shown to serve a critical role in SWM tasks that depend upon the interactions of the HF and mPFC. We demonstrated that the inactivation of RE with muscimol produced deficits in a DNMS T-maze task at short and long delays. These findings are consistent with those of several others reports showing that the disruption of RE produces impairments on various SWM and mnemonic tasks – reportedly dependent on HF-mPFC interactions (Cholvin et al., 2013; Hallock et al.,

2013; Hembrook and Mair, 2011; Hembrook et al., 2012; Layfield et al., 2015; Loureiro et al., 2012, Hallock et al., 2016). For instance, Hallock et al. (2016) demonstrated that the inactivation of RE not only altered SWM performance, but also disrupted theta and gamma synchronous oscillations between the HF and the mPFC during the task. We conclude that RE serves a vital role in SWM in large part by coordinating the actions of the HF and mPFC (Cassel et al., 2013; Vertes et al., 2015b) in delay-dependent tasks.

125

In addition, we showed that the inactivation of RE produced marked perseverative responding, as rats repeatedly made incorrect, non-rewarded choices on T-maze when allowed to correct the responses at no delay time. The deficits we observed are similar to those described in other mnemonic tasks in which the inactivation of RE results in an inability to switch strategies or inhibit inappropriate responses (Cholvin et al., 2013;

Linley et al., 2016). For instance, Cholvin et al. (2013) showed that the inactivation of

RE resulted in the inability of rats to shift from a response (or previously learned) strategy, to a place (or location) strategy in a double H water maze task. In a similar manner, Linley et al. (2016) reported that RE lesions produced deficits in attention and reversal learning (a measure of behavioral flexibility) in an attentional set-shifting task.

We conclude that the inactivation of RE with muscimol disrupts the communication between the hippocampus and the orbitofrontal cortex leading to an inability to alter inappropriate responses (behavioral inflexibility), resulting in a pronounced spatial perseverative behavior.

The impairments observed with the reversible inactivation of RE with muscimol

(Chapter 1) supported a direct role for RE in SWM and executive behaviors. The precise circuitry, however, by which RE exerts these effects on behavior remained to be determined. Accordingly, in Chapter 2, to further assess this, we examined the effects of suppressing RE (directly) or selective RE fibers to the dorsal or ventral HF or the mPFC with an inhibitory DREADD using the same DNMS T-maze task as for Chapter1.

As discussed in Chapter 2, the reversible chemogenetic inactivation of RE impaired SWM performance at both moderate (30s) and long (120s) delays. In addition, the selective suppression of RE projections to the vHF also impaired SWM performance,

126

but only at the moderate (30s) delays. By comparison, the selective suppression of RE projections to the dorsal HF or to the mPFC did not produce significant deficits in SWM performance. Unexpectedly and unlike the effects of muscimol on ‘behavioral flexibility’

(see Chapter 1), inactivation of DREADD infected RE cells or selective RE projections to targets did not produce perseverative behaviors. Finally, the inactivation of RE terminal projections in vHF increased the latency to leave the startbox and also to reach the goal area, whereas RE-dHF inactivation produced the opposite effect: significantly decreasing the latency to leave the start box or reach the goal area. These effects were not related to changes in motor behavior as there were no differences in locomotive speed, but they may reflect differing levels of “anxiety” on the task.

As discussed earlier, the involvement of RE in SWM appears to stem from its role in integrating the actions of the HF and the mPFC. Our finding that chemogenetic inactivation of RE impairs SWM compliments those findings and shows that the more selective inhibition of RE cells with DREADDs produced a very similar effect on SWM as did the pharmacological suppression of RE with muscimol. In addition, we showed that DREADD-induced suppression of RE projections to the ventral HF produced a similar disruption of performance on the DNMS task. Consistent with this, the inactivation of the ventral hippocampus was shown to alter SWM, and also disrupt theta

(O’Neil et al., 2013) and gamma (Spellman et al., 2015) synchronous oscillations between the hippocampus and the mPFC. We conclude that RE is not only critically involved in cognitive (SWM) and executive (behavioral flexibility) processes, but it is also involved in the regulation of affective, motor and attentional actions that are necessary in the proper execution of goal directed behaviors, possibly by influencing and

127

mediating the interactions between, mPFC, OC, dorsal HF and ventral HF during SWM tasks.

In Chapter 3, RE extracellular unit activity from freely behaving rats was recorded as rats cycled through waking-sleep states. We described: (1) a diverse pattern of discharge in RE cells across vigilant states, with most cells firing at high rates in waking and REM sleep and considerably less in SWS; (2) a relatively large population of

RE cells fired maximally during active waking (W1 and W2 cells), with a smaller group discharge maximally in REM sleep (R1 and R2); and (3) most RE cells fired at reduced rates in SWS, with some exhibiting bursting activity in SWS.

To our knowledge, no previous study has described the activity of RE neurons across sleep-waking states. As well documented, the thalamus serves a fundamental role in the modulation of cortical EEG activity during wake and sleep states (Fanselow et at.,

2001; Ward, 2013). Our findings that the majority of RE cells showed increased levels of activity in active W and REM sleep, with reduced rates in SWS, are in line with previous descriptions of ‘specific’ and ‘non-specific’ thalamocortical cells which exhibit a characteristic tonic discharge during wake and REM and less firing during SWS (Steriade et al., 1987; Contreras and Steriade, 1995; Sherman and Guillery, 1996; Sterman, 1996;

Sherman, 2001).

Additionally, we found that a subset of RE cells exhibited a classic bursting pattern of activity during SWS, however, unlike thalamic relay nuclei, the majority of RE cells fired in a tonic fashion during SWS but at significantly reduced rates during SWS.

This observation is consistent with recent studies that described a pattern of reduced

128

firing activity in different thalamic nuclei during SWS (Ramcharan et al., 2000;

Sheroziya and Timofeev, 2014; Liu et al., 2015).

As previously stated (see General Introduction), RE receives considerable input from ‘arousal’ related nuclei of the brainstem such as LC, DR and the LDT/PPT -- involved in the regulation of vigilance states (Moruzzi and Magoun, 1949; Steriade et al,

1993a; Steriade et al, 1993b; Steriade et al 1993c; Sherman and Guillery, 1996). For example, the activity of DR and LC cells is high during W, reduced in SWS and virtually silent in REM sleep, while the discharge of LDT/PPT cells is either selectively high during W or during W and REM sleep (McGinty and Harper, 1976; Aston-Jones and

Bloom, 1981; Carter et al., 2010; Boucetta et al., 2014). The monoaminergic (LC/DR) and cholinergic (LDT/PPT) input to RE could selectively drive subpopulations of RE neurons displaying similar characteristics during vigilance states. Specifically, this would include the “wake-active” and “REM-active” RE neurons.

The changes we observed in RE activity in the transition between vigilant states could suggest that RE, like other thalamic nuclei (Vertes et al., 2015a), could directly modulate cortical EEG activity across vigilance states. This is consistent with the recognized role of RE in arousal/attention (Van der Wef et al., 2002; Vertes et al., 2007).

In this regard, Gent et al. (2018) recently demonstrated that tonic optogenetic stimulation of the midline thalamus in behaving mice during SWS gave rise to behavioral/cortical

EEG activation, whereas a bursting pattern of stimulation produced widespread slow synchronous EEG oscillations in the cortex. Taken together, our unit recording results support a role for RE in arousal, and possibly in sleep-wake state transitions.

129

Summary

Previous studies have demonstrated that RE is critically involved in working memory tasks that depend on the interactions of the HF and the mPFC (Cholvin et al.,

2013; Hallock et al., 2013; Hembrook and Mair, 2011; Hembrook et al., 2012; Loureiro et al., 2012; Hallock et al., 2016). For instance, RE was recently shown to serve a critical role in delayed spatial working memory (SWM) tasks at no or short delays (Layfield et al., 2015; Ito et al., 2015), and to directly participate in executive functions such as strategy-shifting, behavioral inhibition and behavioral flexibility (Dolleman-van der Weel et al., 2009, Cholvin et al., 2013; Prasad et al., 2013; Linley et al., 2016). To further investigate the role of RE in SWM and behavioral flexibility, in Project 1 we examined the effects of reversibly inactivating RE with muscimol on the DNMS T-maze task. We showed that muscimol injections in RE resulted in significant deficits on the task at short and long delays and also importantly produced profound perseverative responses wherein rats repeatedly made incorrect choices on the T-maze in the absence of reward. These findings support a vital role for RE in SWM and executive functions.

Project 2 addressed the possible role of RE projections on each of its major targets and its effect on SWM and behavioral flexibility as compared to the deficits seen in

Project 1. Accordingly, we examined the effects of reversibly inactivating RE or RE terminal projections to the dorsal or ventral hippocampus or to the mPFC with inhibitory

DREADDs using the same DNMS task as in Project 1. We showed that the chemogenetic suppression of RE, or its terminal fibers to the ventral HF, produced significant deficits in SWM performance on the DNMS task. Contrary to our expectations, the selective suppression of RE-dHF or RE-mPFC projections did not result

130

in SWM deficits on the task. Unlike Project 1, we did not observe perseverative behavior for any condition of Project 2. Finally, we found that RE projections to targets differentially affected the times to initiate action and arrive to the goal box (reward) during the sample (encoding) and choice (retrieval) phases of the task: RE-dHF projections accelerated these actions; RE-vHF projections slowed them. Project 2 results supported and extended those of Project 1, further demonstrating a critical role for RE in

SWM, presumably via RE actions on the HF and mPFC.

While some studies have examined the activity of RE cells in behaving rodents with respect to specific tasks or behaviors (Jankowski et al., 2014, 2015; Ito et al., 2015), no previous report of any species had described the discharge properties of RE cells across natural sleep-waking states. In Project 3, we examined the electrophysiological properties of RE neurons across the waking and sleep (SWS, REM) states. We demonstrated that on the whole, most RE neurons discharge in a similar manner to that of other thalamic cells across vigilant states; that is, firing at high rates in waking and REM and at significantly reduced rates in SWS. We, nonetheless, demonstrated a wide diversity among RE neurons as they were classified into five distinct subtypes based on discharge properties across sleep-wake states: W1, W2, R1, R2, and S1 neurons. These various types of RE cells could serve unique and differential roles in arousal/attention, sleep-wake states and cognitive processes (Projects 1 and 2).

In sum, the present work provided new and relevant information regarding the role of RE in arousal-related, cognitive, and executive behaviors. Our findings indicate that RE is critically involved in mnemonic and executive functions, undoubtedly in large

131

part due to its established role in the integration of activity of the hippocampus and the medial prefrontal cortex in cognitive processes.

132

APPENDICES

Appendix A. Published Manuscript Appendix B. Permission to Reproduce Copyrighted Material

133

Appendix A

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

Appendix B

149

REFERENCES

Alexander, G. M., Rogan, S. C., Abbas, A. I., Armbruster, B. N., Pei, Y., Allen, J. A., ... & McNamara, J. O. (2009). Remote control of neuronal activity in transgenic mice expressing evolved G protein-coupled receptors. Neuron, 63(1), 27-39.

Andersen, M. L., Martins, P. J. F., D'almeida, V., Santos, R. F., Bignotto, M., & Tufik, S. (2004). Effects of paradoxical sleep deprivation on blood parameters associated with cardiovascular risk in aged rats. Experimental Gerontology, 39(5), 817-824.

Armbruster, B. N., Li, X., Pausch, M. H., Herlitze, S., & Roth, B. L. (2007). Evolving the lock to fit the key to create a family of G protein-coupled receptors potently activated by an inert ligand. Proceedings of the National Academy of Sciences, 104(12), 5163-5168.

Arthur, D., & Levin, E. D. (2002). Chronic inhibition of α4β2 nicotinic receptors in the ventral hippocampus of rats: impacts on memory and nicotine response. Psychopharmacology, 160(2), 140-145.

Aston-Jones, G., & Bloom, F. E. (1981). Activity of norepinephrine-containing locus coeruleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle. Journal of Neuroscience, 1(8), 876-886.

Aston-Jones, G., & Cohen, J. D. (2005). An integrative theory of locus coeruleus- norepinephrine function: adaptive gain and optimal performance. Annual Review of Neuroscience, 28, 403-450.

Aston-Jones, G., Rajkowski, J., & Cohen, J. (1999). Role of locus coeruleus in attention and behavioral flexibility. Biological Psychiatry, 46(9), 1309-1320.

Baddeley, A. (1986). Working memory. New York, NY, US. Oxford psychology series, No. 11.

Bannerman, D. M., Good, M. A., Butcher, S. P., Ramsay, M., & Morris, R. G. M. (1995). Distinct components of spatial learning revealed by prior training and NMDA receptor blockade. Nature, 378(6553), 182-186.

Benchenane, K., Peyrache, A., Khamassi, M., Tierney, P. L., Gioanni, Y., Battaglia, F. P., & Wiener, S. I. (2010). Coherent theta oscillations and reorganization of spike timing in the hippocampal-prefrontal network upon learning. Neuron, 66(6), 921- 936.

150

Benoit, O., & Chataignier, C. (1973). Patterns of spontaneous unitary discharge in thalamic ventro-basal complex during wakefulness and sleep. Experimental Brain Research, 17(4), 348-363.

Berendse, H. W., & Groenewegen, H. J. (1991). Restricted cortical termination fields of the midline and intralaminar thalamic nuclei in the rat. Neuroscience, 42(1), 73- 102.

Berridge, C. W., Schmeichel, B. E., & España, R. A. (2012). Noradrenergic modulation of wakefulness/arousal. Sleep Medicine Reviews, 16(2), 187-197.

Bertram, E. H., & Zhang, D. X. (1999). Thalamic excitation of hippocampal CA1 neurons: a comparison with the effects of CA3 stimulation. Neuroscience, 92(1), 15-26.

Boucetta, S., Cissé, Y., Mainville, L., Morales, M., & Jones, B. E. (2014). Discharge Profiles across the Sleep–Waking Cycle of Identified Cholinergic, GABAergic, and Glutamatergic Neurons in the Pontomesencephalic Tegmentum of the Rat. The Journal of Neuroscience, 34(13), 4708–4727. doi.org/10.1523/JNEUROSCI.2617-13.2014

Bouret, S., & Sara, S. J. (2005). Network reset: a simplified overarching theory of locus coeruleus noradrenaline function. Trends in Neurosciences, 28(11), 574-582.

Broadbent, N. J., Squire, L. R., & Clark, R. E. (2004). Spatial memory, recognition memory, and the hippocampus. Proceedings of the National Academy of Sciences, 101(40), 14515-14520.

Brown, R. E., Basheer, R., McKenna, J. T., Strecker, R. E., & McCarley, R. W. (2012). Control of sleep and wakefulness. Physiological Reviews, 92(3), 1087-1187.

Brown, R. E., Sergeeva, O. A., Eriksson, K. S., & Haas, H. L. (2002). Convergent excitation of dorsal raphe serotonin neurons by multiple arousal systems (orexin/hypocretin, histamine and noradrenaline). Journal of Neuroscience, 22(20), 8850-8859.

Brown, V. J., & Tait, D. S. (2015). Behavioral flexibility: Attentional shifting, rule switching, and response reversal. Encyclopedia of Psychopharmacology, 264-269.

Buzsáki, G. (1986). Hippocampal sharp waves: their origin and significance. Brain Research, 398(2), 242-252.

Buzsáki, G., & Moser, E. I. (2013). Memory, navigation and theta rhythm in the hippocampal-entorhinal system. Nature Neuroscience, 16(2), 130-135.

Byrne, J. H. (2017). Learning and memory: a comprehensive reference. Academic Press.

151

Campbell, E. J., & Marchant, N. J. (2018). The use of chemogenetics in behavioural neuroscience: receptor variants, targeting approaches and caveats. British Journal of Pharmacology, 175(7), 994-1003.

Canteras, N. S., & Swanson, L. W. (1992). Projections of the ventral subiculum to the amygdala, septum, and hypothalamus: a PHAL anterograde tract‐tracing study in the rat. Journal of Comparative Neurology, 324(2), 180-194.

Carter, M. E., de Lecea, L., & Adamantidis, A. (2013). Functional wiring of hypocretin and LC-NE neurons: implications for arousal. Frontiers in Behavioral Neuroscience, 7, 43. doi: 10.3389/fnbeh.2013.00043.

Carter, M. E., Yizhar, O., Chikahisa, S., Nguyen, H., Adamantidis, A., Nishino, S., ... & De Lecea, L. (2010). Tuning arousal with optogenetic modulation of locus coeruleus neurons. Nature Neuroscience, 13(12), 1526-1533.

Cassel, J. C., De Vasconcelos, A. P., Loureiro, M., Cholvin, T., Dalrymple-Alford, J. C., & Vertes, R. P. (2013). The reuniens and rhomboid nuclei: neuroanatomy, electrophysiological characteristics and behavioral implications. Progress in Neurobiology, 111, 34-52.

Chang, S. E., Todd, T. P., Bucci, D. J., & Smith, K. S. (2015). Chemogenetic manipulation of ventral pallidal neurons impairs acquisition of sign‐tracking in rats. European Journal of Neuroscience, 42(12), 3105-3116.

Cholvin, T., Hok, V., Giorgi, L., Chaillan, F. A., & Poucet, B. (2017). Ventral midline thalamus is necessary for hippocampal place field stability and cell firing modulation. Journal of Neuroscience, 2039-17.

Cholvin, T., Loureiro, M., Cassel, R., Cosquer, B., Geiger, K., Nogueira, D. D. S., ... & Cassel, J. C. (2013). The ventral midline thalamus contributes to strategy shifting in a memory task requiring both prefrontal cortical and hippocampal functions. Journal of Neuroscience, 33(20), 8772-8783.

Churchwell, J. C., & Kesner, R. P. (2011). Hippocampal-prefrontal dynamics in spatial working memory: interactions and independent parallel processing. Behavioural brain research, 225(2), 389-395.

Churchwell, J. C., Morris, A. M., Musso, N. D., & Kesner, R. P. (2010). Prefrontal and hippocampal contributions to encoding and retrieval of spatial memory. Neurobiology of Learning and Memory, 93(3), 415-421.

Cichon, J., & Gan, W. B. (2015). Branch-specific dendritic Ca2+ spikes cause persistent synaptic plasticity. Nature, 520(7546), 180-185.

152

Clark, L., Cools, R., & Robbins, T. W. (2004). The neuropsychology of ventral prefrontal cortex: decision-making and reversal learning. Brain and Cognition, 55(1), 41-53.

Clarke, H. F., Walker, S. C., Crofts, H. S., Dalley, J. W., Robbins, T. W., & Roberts, A. C. (2005). Prefrontal serotonin depletion affects reversal learning but not attentional set shifting. Journal of Neuroscience, 25(2), 532-538.

Clarke, H. F., Walker, S. C., Dalley, J. W., Robbins, T. W., & Roberts, A. C. (2006). Cognitive inflexibility after prefrontal serotonin depletion is behaviorally and neurochemically specific. Cerebral Cortex, 17(1), 18-27.

Cohen, M. S., Kosslyn, S. M., Breiter, H. C., DiGirolamo, G. J., Thompson, W. L., Anderson, A. K., ... & Belliveau, J. W. (1996). Changes in cortical activity during mental rotation. A mapping study using functional MRI. Brain, 119(1), 89-100.

Colgin, L. L. (2011). Oscillations and hippocampal–prefrontal synchrony. Current opinion in neurobiology, 21(3), 467-474.

Contreras, D., & Steriade, M. (1995). Cellular basis of EEG slow rhythms: a study of dynamic corticothalamic relationships. Journal of Neuroscience, 15(1), 604-622.

Cornwall, J., Cooper, J. D., & Phillipson, O. T. (1990). Projections to the rostral reticular thalamic nucleus in the rat. Experimental Brain Research, 80(1), 157-171.

Cornwall, J., & Phillipson, O. T. (1988). Afferent projections to the dorsal thalamus of the rat as shown by retrograde lectin transport. II. The midline nuclei. Brain Research Bulletin, 21(2), 147-161.

Costa-Miserachs, D., Portell-Cortés, I., Torras-Garcia, M., & Morgado-Bernal, I. (2003). Automated sleep staging in rat with a standard spreadsheet. Journal of Neuroscience Methods, 130(1), 93-101.

Coulon, P., Budde, T., & Pape, H. C. (2012). The sleep relay—the role of the thalamus in central and decentral sleep regulation. Pflügers Archiv-European Journal of Physiology, 463(1), 53-71.

Dalley, J. W., Cardinal, R. N., & Robbins, T. W. (2004). Prefrontal executive and cognitive functions in rodents: neural and neurochemical substrates. Neuroscience & Biobehavioral Reviews, 28(7), 771-784.

Darracq, L., Gervasoni, D., Soulière, F., Lin, J. S., Fort, P., Chouvet, G., & Luppi, P. H. (1996). Effect of strychnine on rat locus coeruleus neurones during sleep and wakefulness. Neuroreport, 8(1), 351-355.

Davenne, D. (1987). Lesion of the ponto-geniculo-occipital pathways in kittens. I. Effects on sleep and on unitary discharge of the lateral geniculate nucleus. Brain Research, 409(1), 1-9. 153

Davoodi, F. G., Motamedi, F., Akbari, E., Ghanbarian, E., & Jila, B. (2011). Effect of reversible inactivation of reuniens nucleus on memory processing in passive avoidance task. Behavioural Brain Research, 221(1), 1-6.

Davoodi, F. G., Motamedi, F., Naghdi, N., & Akbari, E. (2009). Effect of reversible inactivation of the reuniens nucleus on spatial learning and memory in rats using Morris water maze task. Behavioural Brain Research, 198(1), 130-135.

Deschênes, M., Paradis, M., Roy, J. P., & Steriade, M. (1984). Electrophysiology of neurons of lateral thalamic nuclei in cat: resting properties and burst discharges. Journal of Neurophysiology, 51(6), 1196-1219.

Dias, R., & Aggleton, J. P. (2000). Effects of selective excitotoxic prefrontal lesions on acquisition of nonmatching‐and matching‐to‐place in the T‐maze in the rat: Differential involvement of the prelimbic–infralimbic and anterior cingulate cortices in providing behavioural flexibility. European Journal of Neuroscience, 12(12), 4457-4466.

Dolleman-Van der Weel, M. J., Da Silva, F. L., & Witter, M. P. (1997). Nucleus reuniens thalami modulates activity in hippocampal field CA1 through excitatory and inhibitory mechanisms. Journal of Neuroscience, 17(14), 5640-5650.

Dolleman-Van der Weel, M. J., Da Silva, F. L., & Witter, M. P. (2017). Interaction of nucleus reuniens and entorhinal cortex projections in hippocampal field CA1 of the rat. Brain Structure and Function, 222(5), 2421-2438.

Dolleman-Van der Weel, M. J., Morris, R. G., & Witter, M. P. (2009). Neurotoxic lesions of the thalamic reuniens or mediodorsal nucleus in rats affect non-mnemonic aspects of watermaze learning. Brain Structure and Function, 213(3), 329-342.

Dormont, J. F. (1972). Patterns of spontaneous unit activity in the ventrolateral thalamic nucleus of cats. Brain Research, 37(2), 223-239.

Dudchenko, P. A. (2001). How do animals actually solve the T maze? Behavioral Neuroscience, 115(4), 850-860.

Dudchenko, P. A. (2004). An overview of the tasks used to test working memory in rodents. Neuroscience & Biobehavioral Reviews, 28(7), 699-709.

Eichenbaum, H. (1999). The hippocampus and mechanisms of declarative memory. Behavioural Brain Research, 103(2), 123-133.

Euston, D. R., Gruber, A. J., & McNaughton, B. L. (2012). The role of medial prefrontal cortex in memory and decision making. Neuron, 76(6), 1057-1070.

154

Fanselow, E. E., Sameshima, K., Baccala, L. A., & Nicolelis, M. A. (2001). Thalamic bursting in rats during different awake behavioral states. Proceedings of the National Academy of Sciences, 98(26), 15330-15335.

Ferbinteanu, J., Ray, C., & McDonald, R. J. (2003). Both dorsal and ventral hippocampus contribute to spatial learning in Long–Evans rats. Neuroscience Letters, 345(2), 131-135.

Floresco, S. B., Seamans, J. K., & Phillips, A. G. (1996). Differential effects of lidocaine infusions into the ventral CA1/subiculum or the nucleus accumbens on the acquisition and retention of spatial information. Behavioural Brain Research, 81(1-2), 163-171.

Floresco, S. B., Seamans, J. K., & Phillips, A. G. (1997). Selective roles for hippocampal, prefrontal cortical, and ventral striatal circuits in radial-arm maze tasks with or without a delay. Journal of Neuroscience, 17(5), 1880-1890.

Foote, S. L., Aston-Jones, G., & Bloom, F. E. (1980). Impulse activity of locus coeruleus neurons in awake rats and monkeys is a function of sensory stimulation and arousal. Proceedings of the National Academy of Sciences, 77(5), 3033-3037.

Frank, L. M., Brown, E. N., & Wilson, M. (2000). Trajectory encoding in the hippocampus and entorhinal cortex. Neuron, 27(1), 169-178.

Ge, F., Wang, N., Cui, C., Li, Y., Liu, Y., Ma, Y., ... & Sun, X. (2017). Glutamatergic projections from the entorhinal cortex to dorsal dentate gyrus mediate context- induced reinstatement of heroin seeking. Neuropsychopharmacology, 42(9), 1860-1870.

Gent, T. C., Bassetti, C. L., & Adamantidis, A. R. (2018). Sleep-wake control and the thalamus. Current Opinion in Neurobiology, 52, 188-197.

Gervasoni, D., Darracq, L., Fort, P., Souliere, F., Chouvet, G., & Luppi, P. H. (1998). Electrophysiological evidence that noradrenergic neurons of the rat locus coeruleus are tonically inhibited by GABA during sleep. European Journal of Neuroscience, 10(3), 964-970.

Gervasoni, D., Peyron, C., Rampon, C., Barbagli, B., Chouvet, G., Urbain, N., ... & Luppi, P. H. (2000). Role and origin of the GABAergic innervation of dorsal raphe serotonergic neurons. Journal of Neuroscience, 20(11), 4217-4225.

Goldman-Rakic, P. S. (1996). Regional and cellular fractionation of working memory. Proceedings of the National Academy of Sciences, 93(24), 13473-13480.

Gomez, J. L., Bonaventura, J., Lesniak, W., Mathews, W. B., Sysa-Shah, P., Rodriguez, L. A., ... & Pomper, M. G. (2017). Chemogenetics revealed: DREADD occupancy and activation via converted clozapine. Science, 357(6350), 503-507. 155

Gordon, J. A. (2011). Oscillations and hippocampal–prefrontal synchrony. Current opinion in neurobiology, 21, 486-491.

Griffin, A. L. (2015). Role of the thalamic nucleus reuniens in mediating interactions between the hippocampus and medial prefrontal cortex during spatial working memory. Frontiers in Systems Neuroscience, 9, 29.

Groenewegen, H. J., & Berendse, H. W. (1994). The specificity of the ‘nonspecific’ midline and intralaminar thalamic nuclei. Trends in Neurosciences, 17(2), 52-57.

Groenewegen, H. J., Berendse, H. W., Wolters, J. G., & Lohman, A. H. (1991). The anatomical relationship of the prefrontal cortex with the striatopallidal system, the thalamus and the amygdala: evidence for a parallel organization. In H.B.M. Uylings, C.G. Van Eden, J.P.C. De Bruin, M.A. Corner, M.G.P. Feenstra (Eds.), Progress in Brain Research (Vol. 85, pp. 95-118). Elsevier.

Groenewegen, H. J., & Witter, M. P. (2004). Thalamus. In Paxinos, G. (Eds.), The Rat Nervous System (Third Edition). Elsevier Academic Press, pp.407–453.

Guettier, J. M., Gautam, D., Scarselli, M., De Azua, I. R., Li, J. H., Rosemond, E., ... & Roth, B. L. (2009). A chemical-genetic approach to study G protein regulation of β cell function in vivo. Proceedings of the National Academy of Sciences, 106(45), 19197-19202.

Gustafson, E. L., Durkin, M. M., Bard, J. A., Zgombick, J., & Branchek, T. A. (1996). A receptor autoradiographic and in situ hybridization analysis of the distribution of the 5‐ht7 receptor in rat brain. British Journal of Pharmacology, 117(4), 657-666.

Halassa, M. M., Siegle, J. H., Ritt, J. T., Ting, J. T., Feng, G., & Moore, C. I. (2011). Selective optical drive of thalamic reticular nucleus generates thalamic bursts and cortical spindles. Nature Neuroscience, 14(9), 1118-1120.

Hallock, H. L., Wang, A., & Griffin, A. L. (2016). Ventral midline thalamus is critical for hippocampal–prefrontal synchrony and spatial working memory. Journal of Neuroscience, 36(32), 8372-8389.

Hallock, H. L., Wang, A., Shaw, C. L., & Griffin, A. L. (2013). Transient inactivation of the thalamic nucleus reuniens and rhomboid nucleus produces deficits of a working-memory dependent tactile-visual conditional discrimination task. Behavioral Neuroscience, 127(6), 860-866.

Hasselmo, M. E. (1999). Neuromodulation: acetylcholine and memory consolidation. Trends in Cognitive Sciences, 3(9), 351-359.

Hembrook, J. R., & Mair, R. G. (2011). Lesions of reuniens and rhomboid thalamic nuclei impair radial maze win‐shift performance. Hippocampus, 21(8), 815-826.

156

Hembrook, J. R., Onos, K. D., & Mair, R. G. (2012). Inactivation of ventral midline thalamus produces selective spatial delayed conditional discrimination impairment in the rat. Hippocampus, 22(4), 853-860.

Herkenham, M. (1978). The connections of the nucleus reuniens thalami: Evidence for a direct thalamo‐hippocampal pathway in the rat. Journal of Comparative Neurology, 177(4), 589-609.

Hobson, J. A., McCarley, R. W., & Wyzinski, P. W. (1975). Sleep cycle oscillation: reciprocal discharge by two brainstem neuronal groups. Science, 189(4196), 55- 58.

Honjoh, S., Sasai, S., Schiereck, S. S., Nagai, H., Tononi, G., & Cirelli, C. (2018). Regulation of cortical activity and arousal by the matrix cells of the ventromedial thalamic nucleus. Nature Communications, 9(1), 2100.

Hoover, W. B., & Vertes, R. P. (2007). Anatomical analysis of afferent projections to the medial prefrontal cortex in the rat. Brain Structure and Function, 212(2), 149- 179.

Hoover, W. B., & Vertes, R. P. (2011). Projections of the medial orbital and ventral orbital cortex in the rat. Journal of Comparative Neurology, 519(18), 3766-3801.

Hoover, W. B., & Vertes, R. P. (2012). Collateral projections from nucleus reuniens of thalamus to hippocampus and medial prefrontal cortex in the rat: a single and double retrograde fluorescent labeling study. Brain Structure and Function, 217(2), 191-209.

Hubel, D. H. (1960). Single unit activity in lateral geniculate body and optic tract of unrestrained cats. The Journal of Physiology, 150(1), 91-104.

Hyman, J. M., Zilli, E. A., Paley, A. M., & Hasselmo, M. E. (2010). Working memory performance correlates with prefrontal-hippocampal theta interactions but not with prefrontal neuron firing rates. Frontiers in Integrative Neuroscience, 4, 2.

Isomura, Y., Sirota, A., Özen, S., Montgomery, S., Mizuseki, K., Henze, D. A., & Buzsáki, G. (2006). Integration and segregation of activity in entorhinal- hippocampal subregions by neocortical slow oscillations. Neuron, 52(5), 871-882.

Ito, H. T., Moser, E. I., & Moser, M.-B. (2018). Supramammillary Nucleus Modulates Spike-Time Coordination in the Prefrontal-Thalamo-Hippocampal Circuit during Navigation. Neuron, (3). doi.org/10.1016/j.neuron.2018.07.021

Ito, H. T., Zhang, S.-J., Witter, M. P., Moser, E. I., & Moser, M.-B. (2015). A prefrontal- thalamo-hippocampal circuit for goal-directed spatial navigation. Nature, (7544), 50. doi.org/10.1038/nature14396.

157

Jahnsen, H., & Llinas, R. (1984). Electrophysiological properties of guinea‐pig thalamic neurones: an in vitro study. The Journal of Physiology, 349(1), 205-226.

Jankowski, M. M., Islam, M. N., Wright, N. F., Vann, S. D., Erichsen, J. T., Aggleton, J. P., & O'Mara, S. M. (2014). Nucleus reuniens of the thalamus contains head direction cells. Elife, 3, e03075.

Jankowski, M. M., Passecker, J., Islam, M. N., Vann, S., Erichsen, J. T., Aggleton, J. P., & O’Mara, S. M. (2015). Evidence for spatially-responsive neurons in the rostral thalamus. Frontiers in Behavioral Neuroscience, 9, 256.

Jay, T. M., & Witter, M. P. (1991). Distribution of hippocampal CA1 and subicular efferents in the prefrontal cortex of the rat studied by means of anterograde transport of Phaseolus vulgaris‐leucoagglutinin. Journal of Comparative Neurology, 313(4), 574-586.

Jinks, A. L., & McGregor, I. S. (1997). Modulation of anxiety-related behaviours following lesions of the prelimbic or infralimbic cortex in the rat. Brain Research, 772(1-2), 181-190.

Jones, B. E., & Moore, R. Y. (1977). Ascending projections of the locus coeruleus in the rat. II. Autoradiographic study. Brain Research, 127(1), 23-53.

Jones, M. W., & Wilson, M. A. (2005). Theta rhythms coordinate hippocampal– prefrontal interactions in a spatial memory task. PLoS Biology, 3(12), e402.

Jung, M. W., Qin, Y., Lee, D., & Mook-Jung, I. (2000). Relationship among discharges of neighboring neurons in the rat prefrontal cortex during spatial working memory tasks. Journal of Neuroscience, 20(16), 6166-6172.

Jung, M. W., Qin, Y., McNaughton, B. L., & Barnes, C. A. (1998). Firing characteristics of deep layer neurons in prefrontal cortex in rats performing spatial working memory tasks. Cerebral Cortex (New York, NY: 1991), 8(5), 437-450.

Jung, M. W., Wiener, S. I., & McNaughton, B. L. (1994). Comparison of spatial firing characteristics of units in dorsal and ventral hippocampus of the rat. Journal of Neuroscience, 14(12), 7347-7356.

Kafetzopoulos, V., Kokras, N., Sotiropoulos, I., Oliveira, J. F., Leite-Almeida, H., Vasalou, A., ... & Sousa, N. (2017). The nucleus reuniens: a key node in the neurocircuitry of stress and depression. Molecular Psychiatry. doi.org/10.1038/mp.2017.55.

Kasamatsu, T., & Adey, W. R. (1973). Visual cortical units associated with phasic activity in REM sleep and wakefulness. Brain Research, 55(2), 323-331.

158

Kjelstrup, K. B., Solstad, T., Brun, V. H., Hafting, T., Leutgeb, S., Witter, M. P., ... & Moser, M. B. (2008). Finite scale of spatial representation in the hippocampus. Science, 321(5885), 140-143.

Kobylarz, E. J., & Schiff, N. D. (2005). Neurophysiological correlates of persistent vegetative and minimally conscious states. Neuropsychological Rehabilitation, 15(3-4), 323-332.

Komorowski, R. W., Garcia, C. G., Wilson, A., Hattori, S., Howard, M. W., & Eichenbaum, H. (2013). Ventral hippocampal neurons are shaped by experience to represent behaviorally relevant contexts. Journal of Neuroscience, 33(18), 8079-8087.

Kraus, B. J., Robinson, R. J., White, J. A., Eichenbaum, H., & Hasselmo, M. E. (2013). Hippocampal “Time Cells”: Time versus Path Integration. Neuron, 78(6), 1090– 1101. doi.org/10.1016/j.neuron.2013.04.015.

Krout, K. E., Belzer, R. E., & Loewy, A. D. (2002). Brainstem projections to midline and intralaminar thalamic nuclei of the rat. Journal of Comparative Neurology, 448(1), 53-101.

Krout, K. E., & Loewy, A. D. (2000). Parabrachial nucleus projections to midline and intralaminar thalamic nuclei of the rat. Journal of Comparative Neurology, 428(3), 475-494.

Lamarre, Y., Filion, M., & Cordeau, J. P. (1971). Neuronal discharges of the ventrolateral nucleus of the thalamus during sleep and wakefulness in the cat I. Spontaneous activity. Experimental Brain Research, 12(5), 480-498.

Lara-Vásquez, A., Espinosa, N., Durán, E., Stockle, M., & Fuentealba, P. (2016). Midline thalamic neurons are differentially engaged during hippocampus network oscillations. Scientific Reports, 6, 29807. . Layfield, D. M., Patel, M., Hallock, H., & Griffin, A. L. (2015). Inactivation of the nucleus reuniens/rhomboid causes a delay-dependent impairment of spatial working memory. Neurobiology of Learning and Memory, 125, 163-167.

Lee, I., & Kesner, R. P. (2003). Time-dependent relationship between the dorsal hippocampus and the prefrontal cortex in spatial memory. Journal of Neuroscience, 23(4), 1517-1523.

Levin, E. D., Bradley, A., Addy, N., & Sigurani, N. (2002). Hippocampal α7 and α4β2 nicotinic receptors and working memory. Neuroscience, 109(4), 757-765.

159

Levin, E. D., Christopher, N. C., Weaver, T., Moore, J., & Brucato, F. (1999). Ventral hippocampal ibotenic acid lesions block chronic nicotine-induced spatial working memory improvement in rats. Cognitive Brain Research, 7(3), 405-410.

Levine, E. S., & Jacobs, B. L. (1992). Neurochemical afferents controlling the activity of serotonergic neurons in the dorsal raphe nucleus: microiontophoretic studies in the awake cat. Journal of Neuroscience, 12(10), 4037-4044.

Lewis, L. D., Voigts, J., Flores, F. J., Schmitt, L. I., Wilson, M. A., Halassa, M. M., & Brown, E. N. (2015). Thalamic reticular nucleus induces fast and local modulation of arousal state. Elife, 4, e08760.

Lichtenberg, N. T., Pennington, Z. T., Holley, S. M., Greenfield, V. Y., Cepeda, C., Levine, M. S., & Wassum, K. M. (2017). Basolateral amygdala to orbitofrontal cortex projections enable cue-triggered reward expectations. Journal of Neuroscience, 0486-17.

Linley, S. B., Gallo, M. M., & Vertes, R. P. (2016). Lesions of the ventral midline thalamus produce deficits in reversal learning and attention on an odor texture set shifting task. Brain Research, 1649, 110-122.

Liu, J., Lee, H. J., Weitz, A. J., Fang, Z., Lin, P., Choy, M., ... & Schiff, N. (2015). Frequency-selective control of cortical and subcortical networks by central thalamus. Elife, 4, e09215.

Liu, T., Bai, W., Xia, M., & Tian, X. (2018). Directional hippocampal-prefrontal interactions during working memory. Behavioural Brain Research, 338, 1-8.

Llinás, R. R., & Steriade, M. (2006). Bursting of thalamic neurons and states of vigilance. Journal of Neurophysiology, 95(6), 3297-3308.

Loureiro, M., Cholvin, T., Lopez, J., Merienne, N., Latreche, A., Cosquer, B., ... & de Vasconcelos, A. P. (2012). The ventral midline thalamus (reuniens and rhomboid nuclei) contributes to the persistence of spatial memory in rats. Journal of Neuroscience, 32(29), 9947-9959.

Loureiro, M., Lecourtier, L., Engeln, M., Lopez, J., Cosquer, B., Geiger, K., ... & De Vasconcelos, A. P. (2012). The ventral hippocampus is necessary for expressing a spatial memory. Brain Structure and Function, 217(1), 93-106.

Mahler, S. V., & Aston-Jones, G. (2018). CNO Evil? Considerations for the use of DREADDs in behavioral neuroscience. Neuropsychopharmacology, 43, 934-936.

160

Mahler, S. V., Vazey, E. M., Beckley, J. T., Keistler, C. R., McGlinchey, E. M., Kaufling, J., ... & Aston-Jones, G. (2014). Designer receptors show role for ventral pallidum input to ventral tegmental area in cocaine seeking. Nature Neuroscience, 17(4), 577.

Marks, G. A., Farber, J., & Roffwarg, H. P. (1981). Phasis influences during REM sleep upon dorsal lateral geniculate nucleus unit activity in the rat. Brain Research, 222(2), 388-394.

Marks, G. A., Roffwarg, H. P., & Shaffery, J. P. (1999). Neuronal activity in the lateral geniculate nucleus associated with ponto-geniculo-occipital waves lacks lamina specificity. Brain Research, 815(1), 21-28.

Marrosu, F., Portas, C., Mascia, M. S., Casu, M. A., Fà, M., Giagheddu, M., ... & Gessa, G. L. (1995). Microdialysis measurement of cortical and hippocampal acetylcholine release during sleep-wake cycle in freely moving cats. Brain Research, 671(2), 329-332.

Martin, J. H., & Ghez, C. (1999). Pharmacological inactivation in the analysis of the central control of movement. Journal of Neuroscience Methods, 86(2), 145-159.

McAlonan, K., & Brown, V. J. (2003). Orbital prefrontal cortex mediates reversal learning and not attentional set shifting in the rat. Behavioural Brain Research, 146(1-2), 97-103.

McCarley, R. W., Benoit, O. D. I. L. E., & Barrionuevo, G. (1983). Lateral geniculate nucleus unitary discharge in sleep and waking: state-and rate-specific aspects. Journal of Neurophysiology, 50(4), 798-818.

McCarley, R. W., Nelson, J. P., & Hobson, J. A. (1978). Ponto-geniculo-occipital (PGO) burst neurons: correlative evidence for neuronal generators of PGO waves. Science, 201(4352), 269-272.

McCormick, D. A., & Pape, H. C. (1990). Properties of a hyperpolarization‐activated cation current and its role in rhythmic oscillation in thalamic relay neurons. The Journal of Physiology, 431(1), 291-318.

McGinty, D. J., & Harper, R. M. (1976). Dorsal raphe neurons: Depression of firing during sleep in cats. Brain Research, 101(3), 569-575.

McGlinchey, E. M., & Aston-Jones, G. (2018). Dorsal hippocampus drives context- induced cocaine seeking via inputs to lateral septum. Neuropsychopharmacology, 43(5), 987.

McKenna, J. T., & Vertes, R. P. (2004). Afferent projections to nucleus reuniens of the thalamus. Journal of Comparative Neurology, 480(2), 115-142.

161

McNaughton, B.L., Barnes, C. A., Battaglia, F., Bower, M., Cowen, S., Ekstrom, A., ... & Lipa, P. (2003). Off-line reprocessing of recent memory and its role in memory consolidation: a progress report. Sleep and Brain Plasticity, 225-246.

Mendelson, W. B., Guthrie, R. D., Frederick, G., & Wyatt, R. J. (1974). The flower pot technique of rapid eye movement (REM) sleep deprivation. Pharmacology Biochemistry and Behavior, 2(4), 553-556.

Miao, C., Cao, Q., Ito, H. T., Yamahachi, H., Witter, M. P., Moser, M. B., & Moser, E. I. (2015). Hippocampal remapping after partial inactivation of the medial entorhinal cortex. Neuron, 88(3), 590-603.

Miller, E. K. (1999). The prefrontal cortex: complex neural properties for complex behavior. Neuron, 22(1), 15–17.

Mölle, M., & Born, J. (2009). Hippocampus whispering in deep sleep to prefrontal cortex—for good memories? Neuron, 61(4), 496-498.

Monti, J. M. (2010a). The role of dorsal raphe nucleus serotonergic and non-serotonergic neurons, and of their receptors, in regulating waking and rapid eye movement (REM) sleep. Sleep Medicine Reviews, 14(5), 319-327.

Monti, J. M. (2010b). The structure of the dorsal raphe nucleus and its relevance to the regulation of sleep and wakefulness. Sleep Medicine Reviews, 14(5), 307-317.

Monti, J. M., & Jantos, H. (2014). The role of serotonin 5-HT7 receptor in regulating sleep and wakefulness. Reviews in the Neurosciences, 25(3), 429-437.

Moore, R. Y., Halaris, A. E., & Jones, B. E. (1978). Serotonin neurons of the midbrain raphe: ascending projections. Journal of Comparative Neurology, 180(3), 417- 438.

Moruzzi, G., & Magoun, H. W. (1949). Brain stem reticular formation and activation of the EEG. Electroencephalography and Clinical Neurophysiology, 1(1-4), 455- 473.

Moser, E., Moser, M. B., & Andersen, P. (1993). Spatial learning impairment parallels the magnitude of dorsal hippocampal lesions, but is hardly present following ventral lesions. Journal of Neuroscience, 13(9), 3916-3925.

Moser, M. B., & Moser, E. I. (1998). Functional differentiation in the hippocampus. Hippocampus, 8(6), 608-619.

Moser, M. B., Moser, E. I., Forrest, E., Andersen, P., & Morris, R. G. (1995). Spatial learning with a minislab in the dorsal hippocampus. Proceedings of the National Academy of Sciences, 92(21), 9697-9701.

162

Munn, N. L. (1950). Handbook of psychological research on the rat; an introduction to animal psychology. Boston: Houghton Mifflin.

Myroshnychenko, M., Seamans, J. K., Phillips, A. G., & Lapish, C. C. (2017). Temporal dynamics of hippocampal and medial prefrontal cortex interactions during the delay period of a working memory-guided foraging task. Cerebral Cortex, 27(11), 5331-5342.

Nitz, D. A., & Siegel, J. M. (1997). Inhibitory amino acid neurotransmission in the dorsal raphe nucleus during sleep/wake states. American Journal of Physiology, 273, R451-R454.

O'Keefe, J., & Conway, D. H. (1978). Hippocampal place units in the freely moving rat: why they fire where they fire. Experimental Brain Research, 31(4), 573-590.

O'Neill, P. K., Gordon, J. A., & Sigurdsson, T. (2013). Theta oscillations in the medial prefrontal cortex are modulated by spatial working memory and synchronize with the hippocampus through its ventral subregion. Journal of Neuroscience, 33(35), 14211-14224.

Ohtake, T., & Yamada, H. (1989). Efferent connections of the nucleus reuniens and the rhomboid nucleus in the rat: an anterograde PHA-L tracing study. Neuroscience Research, 6(6), 556-568.

Padilla-Coreano, N., Bolkan, S. S., Pierce, G. M., Blackman, D. R., Hardin, W. D., Garcia-Garcia, A. L., ... & Gordon, J. A. (2016). Direct ventral hippocampal- prefrontal input is required for anxiety-related neural activity and behavior. Neuron, 89(4), 857-866.

Pare, D., & Steriade, M. (1990). Control of the mammillothalamic axis by brainstem cholinergic laterodorsal tegmental afferents: possible involvement in mnemonic processes. In M. Steriade and D. Biesold (Eds.) Brain Cholinergic Systems, Oxford University Press, Oxford, pp. 337-354.

Pereira de Vasconcelos, A., & Cassel, J. C. (2015). The nonspecific thalamus: A place in a wedding bed for making memories last? Neuroscience & Biobehavioral Reviews, 54, 175-196.

Peschanski, M., & Besson, J. M. (1984). A spino-reticulo-thalamic pathway in the rat: an anatomical study with reference to pain transmission. Neuroscience, 12(1), 165- 178.

Peyrache, A., Khamassi, M., Benchenane, K., Wiener, S. I., & Battaglia, F. P. (2009). Replay of rule-learning related neural patterns in the prefrontal cortex during sleep. Nature Neuroscience, 12(7), 919.

163

Porter, M. C., Burk, J. A., & Mair, R. G. (2000). A comparison of the effects of hippocampal or prefrontal cortical lesions on three versions of delayed non- matching-to-sample based on positional or spatial cues. Behavioural Brain Research, 109, 69–81.

Pothuizen, H. H., Zhang, W. N., Jongen‐Rêlo, A. L., Feldon, J., & Yee, B. K. (2004). Dissociation of function between the dorsal and the ventral hippocampus in spatial learning abilities of the rat: a within‐subject, within‐task comparison of reference and working spatial memory. European Journal of Neuroscience, 19(3), 705-712.

Poucet, B., Thinus-Blanc, C., & Muller, R. U. (1994). Place cells in the ventral hippocampus of rats. Neuroreport: An International Journal for the Rapid Communication of Research in Neuroscience, 5(16), 2045-8.

Prasad, J. A., Abela, A. R., & Chudasama, Y. (2017). Midline thalamic reuniens lesions improve executive behaviors. Neuroscience, 345, 77-88.

Prasad, J. A., Macgregor, E. M., & Chudasama, Y. (2013). Lesions of the thalamic reuniens cause impulsive but not compulsive responses. Brain Structure and Function, 218(1), 85-96.

Preston, A. R., & Eichenbaum, H. (2013). Interplay of hippocampus and prefrontal cortex in memory. Current Biology, 23(17), R764-R773.

Rajkowski, J., Kubiak, P., & Aston-Jones, G. (1994). Locus coeruleus activity in monkey: phasic and tonic changes are associated with altered vigilance. Brain Research Bulletin, 35(5-6), 607-616.

Ramcharan, E. J., Gnadt, J. W., & Sherman, S. M. (2000). Burst and tonic firing in thalamic cells of unanesthetized, behaving monkeys. Visual Neuroscience, 17(1), 55-62.

Rasmussen, K., Morilak, D. A., & Jacobs, B. L. (1986). Single unit activity of locus coeruleus neurons in the freely moving cat: I. During naturalistic behaviors and in response to simple and complex stimuli. Brain Research, 371(2), 324-334.

Rassnick, S., Hoffman, G. E., Rabin, B. S., & Sved, A. F. (1998). Injection of corticotropin-releasing hormone into the locus coeruleus or foot shock increases neuronal Fos expression. Neuroscience, 85(1), 259-268.

Remondes, M., & Wilson, M. A. (2013). Cingulate-hippocampus coherence and trajectory coding in a sequential choice task. Neuron, 80(5), 1277-1289.

Rodgers, C. C., & DeWeese, M. R. (2014). Neural correlates of task switching in prefrontal cortex and primary auditory cortex in a novel stimulus selection task for rodents. Neuron, 82(5), 1157-1170. 164

Rogan, S. C., & Roth, B. L. (2011). Remote control of neuronal signaling. Pharmacological Reviews, 63(2), 291-315.

Rossetti, Z. L., & Carboni, S. (2005). Noradrenaline and dopamine elevations in the rat prefrontal cortex in spatial working memory. Journal of Neuroscience, 25(9), 2322-2329.

Royer, S., Sirota, A., Patel, J., & Buzsáki, G. (2010). Distinct representations and theta dynamics in dorsal and ventral hippocampus. Journal of Neuroscience, 30(5), 1777-1787.

Rushworth, M. F., Noonan, M. P., Boorman, E. D., Walton, M. E., & Behrens, T. E. (2011). Frontal cortex and reward-guided learning and decision-making. Neuron, 70(6), 1054-1069.

Saalmann, Y. B. (2014). Intralaminar and medial thalamic influence on cortical synchrony, information transmission and cognition. Frontiers in Systems Neuroscience, 8, 83.

Santucci, A. C., Knott, P. J., & Haroutunian, V. (1996). Excessive serotonin release, not depletion, leads to memory impairments in rats. European Journal of Pharmacology, 295(1), 7-17.

Sapiurka, M., Squire, L. R., & Clark, R. E. (2016). Distinct roles of hippocampus and medial prefrontal cortex in spatial and nonspatial memory. Hippocampus, 26(12), 1515-1524.

Sawchenko, P. E., & Swanson, L. W. (1983). The organization of forebrain afferents to the paraventricular and supraoptic nuclei of the rat. Journal of Comparative Neurology, 218(2), 121-144.

Schiff, N. D. (2008). Central thalamic contributions to arousal regulation and neurological disorders of consciousness. Annals of the New York Academy of Sciences, 1129(1), 105-118.

Schiff, N. D. (2016). Central thalamic deep brain stimulation to support anterior forebrain mesocircuit function in the severely injured brain. Journal of Neural Transmission, 123(7), 797-806.

Schiff, N. D., Giacino, J. T., Kalmar, K., Victor, J. D., Baker, K., Gerber, M., ... & Farris, S. (2007). Behavioural improvements with thalamic stimulation after severe traumatic brain injury. Nature, 448(7153), 600-603.

Schiff, N. D., & Plum, F. (2000). The role of arousal and “gating” systems in the neurology of impaired consciousness. Journal of Clinical Neurophysiology, 17(5), 438-452.

165

Schiff, N.D., & Purpura, K. (2002). Towards a neurophysiological foundation for cognitive neuromodulation through deep brain stimulation. Thalamus & Related Systems, 2(1), 55-69.

Seamans, J. K., Floresco, S. B., & Phillips, A. G. (1995). Functional differences between the prelimbic and anterior cingulate regions of the rat prefrontal cortex. Behavioral Neuroscience, 109(6), 1063-1073.

Shaw, C., & Aggleton, J. P. (1993). The effects of fornix and medial prefrontal lesions on delayed non-matching-to-sample by rats. Behavioural Brain Research, 54(1), 91- 102.

Sherman, S. M. (2001). Tonic and burst firing: dual modes of thalamocortical relay. Trends in Neurosciences, 24(2), 122-126.

Sherman, S. M., & Guillery, R. W. (1996). Functional organization of thalamocortical relays. Journal of Neurophysiology, 76(3), 1367-1395.

Sherman, S. M., & Guillery, R. W. (2001). Exploring the thalamus. Elsevier.

Sheroziya, M., & Timofeev, I. (2014). Global intracellular slow-wave dynamics of the thalamocortical system. Journal of Neuroscience, 34(26), 8875-8893.

Shima, K., Nakahama, H., & Yamamoto, M. (1986). Firing properties of two types of nucleus raphe dorsalia neurons during the sleep-walking cycle and their responses to sensory stimuli. Brain Research, 399(2), 317-326.

Shin, J. D., & Jadhav, S. P. (2016). Multiple modes of hippocampal–prefrontal interactions in memory-guided behavior. Current opinion in neurobiology, 40, 161-169.

Siapas, A. G., Lubenov, E. V., & Wilson, M. A. (2005). Prefrontal phase locking to hippocampal theta oscillations. Neuron, 46(1), 141-151.

Sigurdsson, T., Stark, K. L., Karayiorgou, M., Gogos, J. A., & Gordon, J. A. (2010). Impaired hippocampal–prefrontal synchrony in a genetic mouse model of schizophrenia. Nature, 464(7289), 763.

Singh, R., Kiloung, J., Singh, S., & Sharma, D. (2008). Effect of paradoxical sleep deprivation on oxidative stress parameters in brain regions of adult and old rats. Biogerontology, 9(3), 153-162.

Sirota, A., Csicsvari, J., Buhl, D., & Buzsáki, G. (2003). Communication between neocortex and hippocampus during sleep in rodents. Proceedings of the National Academy of Sciences, 100(4), 2065-2069.

166

Spellman, T., Rigotti, M., Ahmari, S. E., Fusi, S., Gogos, J. A., & Gordon, J. A. (2015). Hippocampal–prefrontal input supports spatial encoding in working memory. Nature, 522(7556), 309-314.

Sprenger, A., Lappe-Osthege, M., Talamo, S., Gais, S., Kimmig, H., & Helmchen, C. (2010). Eye movements during REM sleep and imagination of visual scenes. Neuroreport, 21(1), 45-49.

Stackman, R. W., Clark, A. S., & Taube, J. S. (2002). Hippocampal spatial representations require vestibular input. Hippocampus, 12(3), 291-303.

Steriade, M. (1996). Arousal--Revisiting the Reticular Activating System. Science, 272(5259), 225-225.

Steriade, M., Contreras, D., Dossi, R. C., & Nunez, A. (1993a). The slow (< 1 Hz) oscillation in reticular thalamic and thalamocortical neurons: scenario of sleep rhythm generation in interacting thalamic and neocortical networks. Journal of Neuroscience, 13(8), 3284-3299.

Steriade, M., Domich, L., Oakson, G., & Deschenes, M. (1987). The deafferented reticular thalamic nucleus generates spindle rhythmicity. Journal of Neurophysiology, 57(1), 260-273.

Steriade, M., & Llinás, R. R. (1988). The functional states of the thalamus and the associated neuronal interplay. Physiological Reviews, 68(3), 649-742.

Steriade, M., McCormick, D. A., & Sejnowski, T. J. (1993b). Thalamocortical oscillations in the sleeping and aroused brain. Science, 262(5134), 679-685.

Steriade, M., Nunez, A., & Amzica, F. (1993c). A novel slow (< 1 Hz) oscillation of neocortical neurons in vivo: depolarizing and hyperpolarizing components. Journal of Neuroscience, 13(8), 3252-3265.

Steriade, M., Paré, D., Bouhassira, D., Deschenes, M., & Oakson, G. (1989). Phasic activation of lateral geniculate and perigeniculate thalamic neurons during sleep with ponto-geniculo-occipital waves. Journal of Neuroscience, 9(7), 2215-2229.

Sterman, M. B. (1996). Physiological origins and functional correlates of EEG rhythmic activities: implications for self-regulation. Biofeedback and Self-regulation, 21(1), 3-33.

Strange, B. A., Witter, M. P., Lein, E. S., & Moser, E. I. (2014). Functional organization of the hippocampal longitudinal axis. Nature Reviews Neuroscience, 15(10), 655.

167

Su, H. S., & Bentivoglio, M. (1990). Thalamic midline cell populations projecting to the nucleus accumbens, amygdala, and hippocampus in the rat. Journal of Comparative Neurology, 297(4), 582-593.

Tamura, M., Spellman, T. J., Rosen, A. M., Gogos, J. A., & Gordon, J. A. (2017). Hippocampal-prefrontal theta-gamma coupling during performance of a spatial working memory task. Nature Communications, 8(1), 2182.

Trulson, M. E., & Jacobs, B. L. (1979). Raphe unit activity in freely moving cats: correlation with level of behavioral arousal. Brain Research, 163(1), 135-150.

Umegaki, H., Munoz, J., Meyer, R. C., Spangler, E. L., Yoshimura, J., Ikari, H., ... & Ingram, D. K. (2001). Involvement of dopamine D2 receptors in complex maze learning and acetylcholine release in ventral hippocampus of rats. Neuroscience, 103(1), 27-33.

Van der Werf, Y. D., Witter, M. P., & Groenewegen, H. J. (2002). The intralaminar and midline nuclei of the thalamus. Anatomical and functional evidence for participation in processes of arousal and awareness. Brain Research Reviews, 39(2-3), 107-140.

Varela, C., Kumar, S., Yang, J. Y., & Wilson, M. A. (2014). Anatomical substrates for direct interactions between hippocampus, medial prefrontal cortex, and the thalamic nucleus reuniens. Brain Structure and Function, 219(3), 911-929.

Vertes, R. P. (1991). A PHA‐L analysis of ascending projections of the dorsal raphe nucleus in the rat. Journal of Comparative Neurology, 313(4), 643-668.

Vertes, R. P. (2002). Analysis of projections from the medial prefrontal cortex to the thalamus in the rat, with emphasis on nucleus reuniens. Journal of Comparative Neurology, 442(2), 163-187.

Vertes, R. P. (2004). Differential projections of the infralimbic and prelimbic cortex in the rat. Synapse, 51(1), 32-58.

Vertes, R. P. (2006). Interactions among the medial prefrontal cortex, hippocampus and midline thalamus in emotional and cognitive processing in the rat. Neuroscience, 142(1), 1-20.

Vertes, R. P., Fortin, W. J., & Crane, A. M. (1999). Projections of the median raphe nucleus in the rat. Journal of Comparative Neurology, 407(4), 555-582.

Vertes, R. P., Hoover, W. B., Do Valle, A. C., Sherman, A., & Rodriguez, J. J. (2006). Efferent projections of reuniens and rhomboid nuclei of the thalamus in the rat. Journal of Comparative Neurology, 499(5), 768-796.

168

Vertes, R. P., Hoover, W. B., Szigeti-Buck, K., & Leranth, C. (2007). Nucleus reuniens of the midline thalamus: link between the medial prefrontal cortex and the hippocampus. Brain Research Bulletin, 71(6), 601-609.

Vertes, R. P., & Linley, S. B. (2008). Efferent and afferent connections of the dorsal and median raphe nuclei in the rat. In Serotonin and Sleep: Molecular, Functional and Clinical Aspects (pp. 69-102). Birkhäuser Basel.

Vertes, R.P., Linley, S.B., Groenewegen, H.J., & Witter, M.P. (2015a). The Thalamus In George Paxinos (Ed.), The rat nervous system (pp. 335-390). Publisher Amsterdam: Elsevier/Academic Press.

Vertes, R. P., Linley, S. B., & Hoover, W. B. (2010). Pattern of distribution of serotonergic fibers to the thalamus of the rat. Brain Structure and Function, 215(1), 1-28.

Vertes, R. P., Linley, S. B., & Hoover, W. B. (2015b). Limbic circuitry of the midline thalamus. Neuroscience & Biobehavioral Reviews, 54, 89-107.

Viana Di Prisco, G., & Vertes, R. P. (2006). Excitatory actions of the ventral midline thalamus (rhomboid/reuniens) on the medial prefrontal cortex in the rat. Synapse, 60(1), 45-55.

Viena, T. D., Linley, S. B., & Vertes, R. P. (2018). Inactivation of nucleus reuniens impairs spatial working memory and behavioral flexibility in the rat. Hippocampus, 28(4), 297-311

Vizuete, M. L., Venero, J. L., Traiffort, E., Vargas, C., Machado, A., & Cano, J. (1997). Expression of 5-HT7 receptor mRNA in rat brain during postnatal development. Neuroscience Letters, 227(1), 53-56.

Walsh, D. A., Brown, J. T., & Randall, A. D. (2017). In vitro characterization of cell‐ level neurophysiological diversity in the rostral nucleus reuniens of adult mice. The Journal of Physiology, 595(11), 3549-3572.

Wang, G. W., & Cai, J. X. (2006). Disconnection of the hippocampal–prefrontal cortical circuits impairs spatial working memory performance in rats. Behavioural Brain Research, 175(2), 329-336.

Wang, G. W., & Cai, J. X. (2008). Reversible disconnection of the hippocampal- prelimbic cortical circuit impairs spatial learning but not passive avoidance learning in rats. Neurobiology of Learning and Memory, 90(2), 365-373.

Ward, L. M. (2013). The thalamus: gateway to the mind. Wiley Interdisciplinary Reviews: Cognitive Science, 4(6), 609-622.

169

Waterhouse, B. D., Border, B., Wahl, L., & Mihailoff, G. A. (1993). Topographic organization of rat locus coeruleus and dorsal raphe nuclei: distribution of cells projecting to visual system structures. Journal of Comparative Neurology, 336(3), 345-361.

Wess, J., Nakajima, K., & Jain, S. (2013). Novel designer receptors to probe GPCR signaling and physiology. Trends in Pharmacological Sciences, 34(7), 385-392.

Weyand, T. G., Boudreaux, M., & Guido, W. (2001). Burst and tonic response modes in thalamic neurons during sleep and wakefulness. Journal of Neurophysiology, 85(3), 1107-1118.

Whissell, P. D., Tohyama, S., & Martin, L. J. (2016). The use of DREADDs to deconstruct behavior. Frontiers in Genetics, 7, 70.

Wikenheiser, A. M., & Schoenbaum, G. (2016). Over the river, through the woods: cognitive maps in the hippocampus and orbitofrontal cortex. Nature Reviews Neuroscience, 17(8), 513.

Witter, M. P., Ostendorf, R. H., & Groenewegen, H. J. (1990). Heterogeneity in the dorsal subiculum of the rat. Distinct neuronal zones project to different cortical and subcortical targets. European Journal of Neuroscience, 2(8), 718-725.

Wood, E. R., Dudchenko, P. A., Robitsek, R. J., & Eichenbaum, H. (2000). Hippocampal neurons encode information about different types of memory episodes occurring in the same location. Neuron, 27(3), 623-633.

Xu, W., & Südhof, T. C. (2013). A neural circuit for memory specificity and generalization. Science, 339(6125), 1290-1295.

Yang, S. T., Shi, Y., Wang, Q., Peng, J. Y., & Li, B. M. (2014). Neuronal representation of working memory in the medial prefrontal cortex of rats. Molecular Brain, 7(1), 61. doi: 10.1186/s13041-014-0061-2.

Yoon, T., Okada, J., Jung, M. W., & Kim, J. J. (2008). Prefrontal cortex and hippocampus subserve different components of working memory in rats. Learning & Memory, 15(3), 97-105.

Young, J. J., & Shapiro, M. L. (2011). Dynamic coding of goal-directed paths by orbital prefrontal cortex. Journal of Neuroscience, 31(16), 5989-6000.

Zhang, Y., Yoshida, T., Katz, D. B., & Lisman, J. E. (2012). NMDAR antagonist action in thalamus imposes delta oscillations on the hippocampus. Journal of Neurophysiology, 107(11), 3181-3189.

170